Search

Lot-sizing and scheduling in a job shop environment is a fundamental problem that appears in many industrial settings. The problem is very complex, and solutions are often needed fast. Although many solution methods have been proposed, with increasingly better results, their computational times are not suitable for decision-makers who want solutions instantly. Therefore, we propose a novel greedy heuristic to efficiently generate production plans and schedules of good quality. The main innovation of our approach represents the incorporation of a simulation-based technique, which directly generates schedules while simultaneously determining lot sizes. By utilising priority rules, this unique feature enables us to address the complexity of job shop scheduling environments and ensures the feasibility of the resulting schedules. Using a selection of well-known rules from the literature, experiments on a variety of shop configurations and complexities showed that the proposed heuristic is able to obtain solutions with an average gap to Cplex of 4.12%. To further improve the proposed heuristic, a cooperative coevolutionary genetic programming-based hyper-heuristic has been developed. The average gap to Cplex was reduced up to 1.92%. These solutions are generated in a small fraction of a second, regardless of the size of the instance. [ABSTRACT FROM AUTHOR]

Genetic programming has achieved great success in project scheduling by generating Priority Rules (PRs) through evolution. However, the frequent disturbance factors in practice not only lead to the appropriate PR changes in different states, but also increase the calculation consumption in evaluation. In this paper, a novel Hyper-heuristic-based Surrogate-Assisted Dual-Tree Genetic Programming (HSDGP) framework is proposed for the Dynamic Resource Constrained Multi-Project Scheduling Problem with new project Insertions and resource Disruptions (DRCMPSP-ID). Uniquely, the proposed method automatically evolves two PRs for scheduling DRCMPSP-ID under normal and disruptive states respectively, and an expansion search mechanism based on neighbourhood is designed to improve PR generation ability by generating a large number of offspring and implement the search of dual-tree encoding. Furthermore, in order to estimate the fitness of new individuals, an activity-sequence based surrogate is proposed to deal with the input of activity sequence during schedule generation and reduce the evaluation calculation consumption. Based on the instances constructed by the existing benchmark, the experimental result shows the superiority of HSDGP and the impact of key parameters on its performance. [ABSTRACT FROM AUTHOR]

The AUTO algorithm is presented to incrementally build models and solve supervised learning problems. AUTO uses traditional derivative-free optimisation, like genetic algorithms, to search the problem space of arbitrary functions. With reinforcement learning, AUTO learns actions to guide its search process and gradually improve performance. A comparative analysis is presented exploring supervised learning performance and interpretability of AUTO models. Results indicate AUTO outperforms genetic programming and rivals multilayer perceptrons. AUTO automatically builds models to closely match datasets without the limited search space of traditional supervised learning. With enough time and computational resources, AUTO can generate any function. [ABSTRACT FROM AUTHOR]

Large variations across images make image classification a challenging task; limited training data further increases its difficulty. Genetic programming (GP) has been considerably applied to image classification. However, most GP methods tend to directly evolve a single classifier or depend on a predefined classification algorithm, which typically does not lead to ideal generalization performance when only a few training instances are available. Applying ensemble learning to classification often outperforms employing a single classifier. However, single-tree representation (each individual contains a single tree) is widely employed in GP. Training multiple diverse and accurate base learners/classifiers based on single-tree GP is challenging. Therefore, this article proposes a new ensemble construction method based on multi-tree GP (each individual contains multiple trees) for image classification. A single individual forms an ensemble, and its multiple trees constitute base learners. To find the best individual in which multiple trees are diverse and effectively cooperate, i.e., the nth tree can correct the errors of the previous n-1 trees, the new method assigns different weights to multiple trees using the idea of AdaBoost and performs classification via weighted majority voting. Furthermore, a new tree representation is developed to evolve diverse and accurate base learners that extract useful features and conduct classification simultaneously. The new approach achieves significantly better performance than almost all benchmark methods on eight datasets. Additional analyses highlight the effectiveness of the new ensembles and tree representation, demonstrating the potential for providing valuable interpretability in ensemble trees. [ABSTRACT FROM AUTHOR]

The presented work outlines a new genetic crossover operator, which can be used to solve problems by the Grammatical Evolution technique. This new operator intensively applies the one-point crossover procedure to randomly selected chromosomes with the aim of drastically reducing their fitness value. The new operator is applied to chromosomes selected randomly from the genetic population. This new operator was applied to two techniques from the recent literature that exploit Grammatical Evolution: artificial neural network construction and rule construction. In both case studies, an extensive set of classification problems and data-fitting problems were incorporated to estimate the effectiveness of the proposed genetic operator. The proposed operator significantly reduced both the classification error on the classification datasets and the feature learning error on the fitting datasets compared to other machine learning techniques and also to the original models before applying the new operator. [ABSTRACT FROM AUTHOR]

Constructed neural networks with the assistance of grammatical evolution have been widely used in a series of classification and data-fitting problems recently. Application areas of this innovative machine learning technique include solving differential equations, autism screening, and measuring motor function in Parkinson's disease. Although this technique has given excellent results, in many cases, it is trapped in local minimum and cannot perform satisfactorily in many problems. For this purpose, it is considered necessary to find techniques to avoid local minima, and one technique is the periodic application of local minimization techniques that will adjust the parameters of the constructed artificial neural network while maintaining the already existing architecture created by grammatical evolution. The periodic application of local minimization techniques has shown a significant reduction in both classification and data-fitting problems found in the relevant literature. [ABSTRACT FROM AUTHOR]

Several constrained optimization problems have been adequately solved over the years thanks to the advances in the area of metaheuristics. Nevertheless, the question as to which search logic performs better on constrained optimization often arises. In this paper, we present Dual Search Optimization (DSO), a co-evolutionary algorithm that includes an adaptive penalty function to handle constrained problems. Compared to other self-adaptive metaheuristics, one of the main advantages of DSO is that it is able auto-construct its own perturbation logics, i.e., the ways solutions are modified to create new ones during the optimization process. This is accomplished by co-evolving the solutions (encoded as vectors of integer/real values) and perturbation strategies (encoded as Genetic Programming trees), in order to adapt the search to the problem. In addition to that, the adaptive penalty function allows the algorithm to handle constraints very effectively, yet with a minor additional algorithmic overhead. We compare DSO with several algorithms from the state-of-the-art on two sets of problems, namely: (1) seven well-known constrained engineering design problems and (2) the CEC 2017 benchmark for constrained optimization. Our results show that DSO can achieve state-of-the-art performances, being capable to automatically adjust its behavior to the problem at hand. [ABSTRACT FROM AUTHOR]

Symbolic regression plays a crucial role in machine learning and data science by allowing the extraction of meaningful mathematical models directly from data without imposing a specific structure. This level of adaptability is especially beneficial in scientific and engineering fields, where comprehending and articulating the underlying data relationships is just as important as making accurate predictions. Genetic Programming (GP) has been extensively utilized for symbolic regression and has demonstrated remarkable success in diverse domains. However, GP's heavy reliance on evolutionary mechanisms makes it computationally intensive and challenging to handle. On the other hand, Particle Swarm Optimization (PSO) has demonstrated remarkable performance in numerical optimization with parallelism, simplicity, and rapid convergence. These attributes position PSO as a compelling option for Automatic Programming (AP), which focuses on the automatic generation of programs or mathematical models. Particle Swarm Programming (PSP) has emerged as an alternative to Genetic Programming (GP), with a specific emphasis on harnessing the efficiency of PSO for symbolic regression. However, PSP remains unsolved due to the high-dimensional search spaces and local optimal regions in AP, where traditional PSO can encounter issues such as premature convergence and stagnation. To tackle these challenges, we introduce Dynamical Sphere Regrouping PSO Programming (DSRegPSOP), an innovative PSP implementation that integrates DSRegPSO's dynamical sphere regrouping and momentum conservation mechanisms. DSRegPSOP is specifically developed to deal with large-scale, high-dimensional search spaces featuring numerous local optima, thus proving effective behavior for symbolic regression tasks. We assess DSRegPSOP by generating 10 mathematical expressions for mapping points from functions with varying complexity, including noise in position and cost evaluation. Moreover, we also evaluate its performance using real-world datasets. Our results show that DSRegPSOP effectively addresses the shortcomings of PSO in PSP by producing mathematical models entirely generated by AP that achieve accuracy similar to other machine learning algorithms optimized for regression tasks involving numerical structures. Additionally, DSRegPSOP combines the benefits of symbolic regression with the efficiency of PSO. [ABSTRACT FROM AUTHOR]

Accurate assessment of scour depth around bridge abutments is crucial to reasonable design of abutment structures. In this study, machine learning (ML) models are implemented, including M5′ model tree (M5′MT), multivariate adaptive regression spline (MARS), locally weighted polynomial regression (LWPR) and multigene genetic programming (MGGP) to predict scour depth around vertical-wall, 45° wing-wall and semicircular bridge abutments. Published experimental data are adopted, with four input parameters considered for the prediction of relative scour depth. The optimal input combination for each model is first determined using correlation and sensitivity analyses; results reveal that MGGP exhibits the best agreement with experimental data for vertical-wall and semicircular abutments, whereas LWPR outperforms the other models for the 45° wing-wall abutment. In addition, compared with the empirical equations and ML models employed in the literature, the accuracy of scour depth prediction is significantly improved with the ML models used in this study. Considering the comprehensive performance for all types of abutments in terms of accuracy, reliability and interpretability, MGGP is recommended as the representative of the implemented ML models with its mean absolute percentage error of 2.40% for a vertical-wall abutment, 3.95% for a 45° wing-wall abutment and 3.85% for a semicircular abutment. [ABSTRACT FROM AUTHOR]

BACKGROUND: Strawberry (Fragaria×ananassa Duch.) is a species of great economic and nutritional importance. It is widely cultivated in different regions of the world. However, climatic factors have a significant influence on its production. OBJECTIVE: To evaluate the ecoclimatic suitability and growth index of F.×ananassa in the face of the impacts of climate change in Brazil and the world. METHODS: The potential global distribution of F.×ananassa it was based on predicted global climate changes using CLIMEX and the A1B and A2 emissions scenario for the years 2050 and 2100. CLIMEX weekly growth index (GIw) was applied, and productivity data in commercial strawberry cultivation in Brazilian areas were collected. RESULTS: The results indicate a progressive reduction in areas suitable for strawberry cultivation globally, mainly on the African continent, Australia, China, and the Americas for the years 2050 and 2100. In Brazil, only the South region and part of the Southeast will continue to be suitable for strawberry production. CONCLUSIONS: The CLIMEX model predicts a reduction in the suitable area for cultivating Fragaria×ananassa Duch between 2050 and 2100 due to climate change. The generated maps can help identify new cultivation areas and support genetic improvement programmes to develop cultivars more tolerant to heat. [ABSTRACT FROM AUTHOR]

Accurate machine learning streamflow prediction often requires coupling data-driven models with preprocessing techniques. This study aims to improve the performance of deep learning (DL) models, including long short-term memory, recurrent neural network (RNN), and gated recurrent unit (GRU) by incorporating maximal overlap discrete wavelet entropy transform (MODWET) techniques for streamflow forecasting. The merit of MODWET over maximal overlap discrete wavelet transform (MODWT) is that MODWET utilizes Entropy to determine the optimal decomposition level and suitable wavelet function, which was an unaddressed problem in wavelet-based decomposition models. Suitable decomposition level prevents providing unnecessary information or missing essential information. In this study we show that a unique decomposition level and wavelet filter is not suitable for any dataset. The research focuses on monthly streamflow data from three case studies in the CAMEL dataset in the USA. The accuracy of the models is evaluated using statistical measures such as Nash–Sutcliffe efficiency (NSE), root-mean-squared error, percent bias, and correlation coefficient (r). To determine the optimal model, a Taylor diagram is utilized. The results demonstrate the effectiveness of coupling MODWET with DL models in flood forecasting. Furthermore, genetic programming (GP) and partial correlation index (PCI) are employed for predictor selection. Hybrid models, namely MODWET-GP-GRU (NSE of 0.83), MODWET-GP-RNN (NSE of 0.95), and MODWET-PCI-GRU (NSE of 0.95), outperform simple DL models in terms of NSE and Taylor diagram evaluation. This study emphasizes the potential of hybrid models that combine DL algorithms with the recently proposed MODWET technique for streamflow prediction. [ABSTRACT FROM AUTHOR]

Many real-world problems have an uneven distribution of data over different classes. The imbalanced nature of data impacts the performance of classifiers. The higher counts of majority class samples influence the learning abilities of well-known classifiers. Genetic programming (GP) algorithm based on natural evolution also impacts if the data's nature is imbalanced. The fitness function plays a pivotal role and impacts almost each building block of the GP framework. GP with the standard fitness function produces under-fitted and biased classifiers. Therefore, this paper has proposed a new fitness function in GP to classify the imbalanced data. The proposed method is used to classify nine imbalanced problems: ABL-18, ABL-9-18, BAL, YEAST2, YEAST1, ABL-9, ION, WDBC, and SPECT. The imbalanced factor of benchmark problems varies from 99:1 to 59:41. The proposed method's performance is compared with K-Nearest-Neighbourhood (KNN) and the standard fitness function-based GP methods. The GP with newly proposed fitness function gives average AUC values for ABL-18(99:1), ABL-9-18(94:6), BAL(92:8), YEAST2(89:11), YEAST1(84:16), ABL-9(83:17), ION(64:36), WDBC(63:37), and SPECT(59:41) as 0.714, 0.812, 0.975, 0.916, 0.768, 0.654, 0.872, 0.939, and 0.704, respectively, which are higher than KNN and the standard fitness function-based GP methods. The result outcomes prove the superiority of the proposed method. [ABSTRACT FROM AUTHOR]

One of the big challenges in machine learning is difficulty of achieving high accuracy in a short completion time. A more difficulties appeared when the algorithm needs to be used for solving real dataset from the survey-based data collection. Imbalance dataset, insufficient strength of correlations, and outliers are common problems in real dataset. To accelerate the modelling processes, automated machine learning based on meta-heuristics optimization such as genetic programming (GP) has started to emerge and is gaining popularity. However, identifying the best hyper-parameters of the meta-heuristics' algorithm is the critical issue. This paper demonstrates the evaluation of GP hyper-parameters in modeling machine learning on house affordability dataset. The important hyper-parameters of GP are population size (PS), that has been observed with different setting in this research. The machine learning with GP was used to predict house affordability among employers with transport expenditure and job mobility as some of the attributes. The results from testing that run on hold-out samples show that GP machine learning can reach to 70% accuracy with split ratio 0.2 and GP PS 30. This research contributes to the advancement of automated machine learning techniques, offering potential for faster and more accurate real survey-based datasets. [ABSTRACT FROM AUTHOR]

Process mining algorithms essentially reflect the execution behavior of events in an event log for conformance checking, model discovery, or enhancement. Domain experts have developed several process mining algorithms based on theoretical frameworks such as linear integer programming, heuristics, and genetic algorithms, region-based and semantic-based approaches. The idea is to generate insightful representations of these processes of information systems to enable process mining practitioners to gain insight into their systems. Recently, there has been a shift toward semantic-based approaches for process mining since they not only discover enhanced models but also emphasize context. To this effect, this paper conducts a comprehensive review of 30 articles on semantic process mining techniques. It was found that 44.7% of all works used semantics for process discovery, 23.7% for model enhancement, and conformance checking was the least with 10.5%. We further indicate the benefits and contributions of these methods to process mining. Challenges, opportunities, and prospective future research areas are also discussed. [ABSTRACT FROM AUTHOR]

Due to the explosive growth of software quantity and the mixed ability of software developers, a large number of software defects emerge during the later stages of software maintenance. The search method based on genetic programming is one of the most popular in search algorithms, but it also has some issues. The single-objective approach to validate and select offspring patches without considering other constraints can affect the efficiency of patch generation. To address this issue, this paper proposes an automatic software repair method based on Multi-objective Genetic Programming (MGPRepair). Firstly, the method adopts a lightweight context analysis strategy to find suitable repair materials. Secondly, it decouples the replacement statements and insertion statements in the repair materials, using a lower-granularity patch representation method to encode the patches in the search space. Then, the automatic software defect repair is treated as a multi-objective search problem, and the NSGA-II multi-objective optimization algorithm is used to find simpler repair patches. Finally, the test case filtering technique is used to accelerate the patch validation process and generate correct patches. MGPRepair was experimentally evaluated on 395 real Java software defects from the Defects4J dataset. The experimental results show that MGPRepair can generate test case-passing patches for 51 defects, of which 35 defect patches are equivalent to manually generated patches. Its repair the efficiency and success rate are higher to other excellent automatic software defect repair methods such as jGenProg, RSRepair, ARJA, Nopol, Capgen, and SequenceR. [ABSTRACT FROM AUTHOR]

Portland cement concrete (PCC) is a major contributor to human-made CO2 emissions. To address this environmental impact, fly ash geopolymer concrete (FAGC) has emerged as a promising low-carbon alternative. This study establishes a robust compressive strength prediction model for FAGC and develops an optimal mixture design method to achieve target compressive strength with minimal CO2 emissions. To develop robust prediction models, comprehensive factors, including fly ash characteristics, mixture proportions, curing parameters, and specimen types, are considered, a large dataset comprising 1136 observations is created, and polynomial regression, genetic programming, and ensemble learning are employed. The ensemble learning model shows superior accuracy and generalization ability with an RMSE value of 1.81 MPa and an R2 value of 0.93 in the experimental validation set. Then, the study integrates the developed strength model with a life cycle assessment-based CO2 emissions model, formulating an optimal FAGC mixture design program. A case study validates the effectiveness of this program, demonstrating a 16.7% reduction in CO2 emissions for FAGC with a compressive strength of 50 MPa compared to traditional trial-and-error design. Moreover, compared to PCC, the developed FAGC achieves a substantial 60.3% reduction in CO2 emissions. This work provides engineers with tools for compressive strength prediction and low carbon optimization of FAGC, enabling rapid and highly accurate design of concrete with lower CO2 emissions and greater sustainability. [ABSTRACT FROM AUTHOR]

The development of concrete with excellent water and frost resistance providing high level of sound and thermal insulation has triggered the formulation of foamed concrete. However, multiple laboratory studies are required to produce reasonable data to design the relevant codes and mathematics with which design of mixes is made easier at low cost. In this research paper, the artificial intelligence (AI)-based symbolic regression technique estimation of the compressive strength of foamed concrete has been reported. Foamed concrete has been a subject of serious research in sustainable built-environment due to its lightweight and structural functionality. In this research work, data gathering method was applied to gather a globally representative data base comprising concrete density to water density (concrete density g/cm3) (y/yw), water-cement ratio (W/C), and sand-cement ratio (S/C) as input variable and the compressive strength (Fc) as the study output. The dimensionless factors have been derived to eliminate data handling complexities and improve model performances. The 230 data entries from foamed concrete mixes were partitioned into 75% and 25% for training and validation data, respectively. At the end of the model execution, it was found that the response surface methodology (RSM) produced a symbolic closed-form equation like the genetic programming (GP), evolutionary polynomial regression (EPR), and the group method of data-handling-neural network (GMDH-NN). Even though the RSM closed with a minimum error, the GP, EPR and GMDH-NN were faster in runtime. The overall outcomes show that the GP outclassed the EPR, RSM and the GMDH-NN, though with minor margin. Meanwhile the EPR produced the highest outliers from the ±25% test of accuracy envelope. Overall, the present models outperformed those reported in the literature due the parameter reduction through dimensionless factors derivation and provided a decisive model to predict the Fc of foamed concrete. [ABSTRACT FROM AUTHOR]

Distributed processing and assembly multistage manufacturing system consist of multiple job shop for processing parts and a general assembly flow shop for partial and final assembly. Dynamic orders with multi-layer product structures require part matching before assembly. The control of such multistage shop involves joint decision-making on order allocation, processing, and assembly jobs scheduling for precise coordination between the two levels of production. An automatic design method of scheduling rules for distributed machining and assembly multistage workshops based on genetic programming based on parallel simulation(GP-PS) was proposed. Firstly, a mathematical model was established to minimize the percentage of tardy product. Then, an improved genetic programming algorithm was introduced to integrate evolutionary multistage scheduling rules. A population optimization mechanism was designed to prevent the algorithm from converging locally, while parallel simulation technology was embedded to effectively reduce training time. Finally, simulation experiments were conducted to validate the performance of the improved genetic algorithm. The experimental results show that the percentage of tardy product obtained from the manual rule set, standard genetic programming, and improved algorithm is respectively 6. 44%, 5. 65% and 2. 67%. The improved GP algorithm based on parallel simulation optimization can achieve more obvious comprehensive performance advantages than dozens of preferred artificial rule groups and the optimal rule groups generated by standard GP algorithm. It is feasible and effective to use this algorithm to automatically design multi-level rules of integrated scheduling for DPAMW scheduling problem. [ABSTRACT FROM AUTHOR]

Genetic programming (GP) has a long-standing tradition in the evolution of computer programs, predominantly utilizing tree and linear paradigms, each with distinct advantages and limitations. Despite the rapid growth of the GP field, there have been disproportionately few attempts to evolve 'real' Turing-like imperative programs (as contrasted with functional programming) from the ground up. Existing research focuses mainly on specific special cases where the structure of the solution is partly known. This paper explores the potential of integrating tree and linear GP paradigms to develop an encoding scheme that universally supports genetic operators without constraints and consistently generates syntactically correct Python programs from scratch. By blending the symmetrical structure of tree-based representations with the inherent asymmetry of linear sequences, we created a versatile environment for program evolution. Our approach was rigorously tested on 35 problems characterized by varying Halstead complexity metrics, to delineate the approach's boundaries. While expected brute-force program solutions were observed, our method yielded more sophisticated strategies, such as optimizing a program by restricting the division trials to the values up to the square root of the number when counting its proper divisors. Despite the recent groundbreaking advancements in large language models, we assert that the GP field warrants continued research. GP embodies a fundamentally different computational paradigm, crucial for advancing our understanding of natural evolutionary processes. [ABSTRACT FROM AUTHOR]

In this work we aim to empirically characterize two important dynamical aspects of GP search: the evolution of diversity and the propagation of inheritance patterns. Diversity is calculated at the genotypic and phenotypic levels using efficient similarity metrics. Inheritance information is obtained via a full genealogical record of evolution as a directed acyclic graph and a set of methods for extracting relevant patterns. Advances in processing power enable our approach to handle previously infeasible graph sizes of millions of arcs and vertices. To enable a more comprehensive analysis we employ three closely-related but different evolutionary models: canonical GP, offspring selection and age-layered population structure. Our analysis reveals that a relatively small number of ancestors are responsible for producing the majority of descendants in later generations, leading to diversity loss. We show empirically across a selection of five benchmark problems that each configuration is characterized by different rates of diversity loss and different inheritance patterns, in support of the idea that each new problem may require a unique approach to solve optimally. [ABSTRACT FROM AUTHOR]

Robust segmentation of large and complex conjoined tree structures in 3-D is a major challenge in computer vision. This is particularly true in computational biology, where we often encounter large data structures in size, but few in number, which poses a hard problem for learning algorithms. We show that merging multiscale opening with geodesic path propagation, can shed new light on this classic machine vision challenge, while circumventing the learning issue by developing an unsupervised visual geometry approach (digital topology/morphometry). The novelty of the proposed MSO-GP method comes from the geodesic path propagation being guided by a skeletonization of the conjoined structure that helps to achieve robust segmentation results in a particularly challenging task in this area, that of artery-vein separation from non-contrast pulmonary computed tomography angiograms. This is an important first step in measuring vascular geometry to then diagnose pulmonary diseases and to develop image-based phenotypes. We first present proof-of-concept results on synthetic data, and then verify the performance on pig lung and human lung data with less segmentation time and user intervention needs than those of the competing methods. • We revisit the classic computer vision problem of segmenting complex conjoined tree-like structures. • We the branches as skeleton guided geodesic paths and open them morphologically in multiple scales. • The proposed MSO-GP method is unsupervised and non-learning based, hence fast and annotation independent. • We demonstrate the method on artery-vein segmentation in non contrast lung computed tomography (CT) angiograms. • MSO-GP outperforms two competing methods on synthetic generated data, pig lung phantom data and human lung data. [ABSTRACT FROM AUTHOR]

This study considers multiperiod bankruptcy prediction models, an aspect scarcely considered in research despite its importance, since creditors must assess the risk of loans over the entire life of the debt and not at a specific point in the future. Two possibilities for the implementation of multiperiod prediction models are considered: Multi-Model multiperiod Bankruptcy Prediction Models (MMBPM) and Single-Model multiperiod Bankruptcy Prediction Models (SMBPM). The former considers the conditional probabilities obtained by individual models predicting bankruptcy at specific times in the future, while the latter is a single model predicting bankruptcy at a specific time interval in the future. The results show that there are no significant differences between the two approaches when compared using data after the learning period. However, SMBPMs have the important advantage of interpretability for decision-making, which is discussed with examples. Moreover, a comparison of SMBPM performance with external references is performed. [ABSTRACT FROM AUTHOR]

Over the past decade, the surge in global container port throughput has heightened the demand for terminal efficiency, with the container yard operations being central to the overall port performance. However, the unpredictable arrival of external trucks poses significant challenges for yard cranes which must simultaneously schedule operations for both internal and external tasks. Traditional yard crane scheduling methods often rely on outdated assumptions that fail to account for the dynamic impact of external tasks. In response, container terminals increasingly model the yard crane scheduling as an online problem. A notable advancement in online scheduling is the online rollout method, which evaluates the decisions based on the potential outcomes of their future rollout schedules rather than immediate priorities. Although this method outperforms the previous approach, it faces two main issues: the rollout simulation is time consuming, and decisions based solely on objective value of rollout schedules may not align with long-term scheduling objectives. To overcome these limitations, we have developed a two-stage adaptive rollout decision model. In the first stage, less desirable tasks are dynamically filtered out to reduce the number of rollout simulations required, while the second stage employs a genetic programming evolved evaluation function to infuse more refined forward-looking insights into the scheduling process. This approach has proven to significantly enhance yard scheduling efficiency and performance, as confirmed by experimental validation. Given the dynamic nature of yard crane operations, we believe this method could be beneficially applied to other dynamic scheduling contexts. [ABSTRACT FROM AUTHOR]

A key issue for effective management and operating of dam reservoirs is predicting the water inflow values into dam reservoir. To address this subject, here, genetic programming (GP) is used by proposing two cases. In the first case, water inflow values are predicted separately for each month. However, in the second case, these values are predicted simultaneously for all months. Furthermore, for each case, two approaches are proposed here. In the first approach, the hybrid method, called the ANN-NGSA-II method, is proposed to find proper input data sets. However, in the second approach, the useful input data sets are found automatically using the GP method. For comparison purpose, the ANN and SARIMA models are also used, to predict the water inflow values. As a case study, in this research, the Zayandehroud dam reservoir is selected. The results indicate that the ANN model outperforms both results of the GP and SARIMA methods. In other words, correlation coefficient (R2), Nash Sutcliffe (NS), and root means square error (RMSE) values of ANN are 0.97, (0.88), 0.954 (0.87), and 17.19 (30.54) million cubic meters, respectively, for training (test) data set. [ABSTRACT FROM AUTHOR]

This study focuses on the deterioration of the predictive power and the analysis of the predictive stability of business failure prediction models, an aspect not sufficiently analysed in previous research. Insolvency prediction is considered with three temporal horizons (1 year, 3 years and 5 years prior to failure). The Genetic Programming (GP) tool has been used to achieve prediction models with high performance and stability over time, considering a long post-learning period in the stability analysis. In addition, novel scenarios representative of actual model use are proposed and considered, as well as metrics to assess the deterioration of the models’ predictive power. The optimised GP prediction models (in the three temporal horizons) present a higher performance with respect to external references and, more importantly in relation to the objective of our study, the selected GP models substantially improve on the stability reported in previous studies, meeting the pursued requirements of degree of deterioration (less than 5%) and stability (Pearson’s coefficient of variation less than 5%). Thus, the predictions of the GP models after the learning are very stable (period 2008–2019), to a certain extent immune, with respect to their environment, responding adequately in both procyclical and countercyclical modes, all of which is particularly relevant as this period includes a strong recession and a strong recovery. This should help to increase the reliability of business failure prediction models. Moreover, the relevance of including variables other than the usual financial ratios as predictors of failure is confirmed. [ABSTRACT FROM AUTHOR]

The classification of Software-Defined Networking (SDN) traffic is an essential tool for network management, network monitoring, traffic engineering, dynamic resource allocation planning, and applying Quality of Service (QoS) policies. The programmability nature of SDN, the holistic view of the network through SDN controllers, and the capability for dynamic adjustable and reconfigurable controllersare fertile ground for the development of new techniques for traffic classification. Although there are enough research works that have studied traffic classification methods in SDN environments, they have several shortcomings and gaps that need to be further investigated. In this study, we investigated traffic classification methods in SDN using publicly available SDN traffic trace datasets. We apply a series of classifiers, such as MLP (BFGS), FC2 (RBF), FC2 (MLP), Decision Tree, SVM, and GENCLASS, and evaluate their performance in terms of accuracy, detection rate, and precision. Of the methods used, GenClass appears to be more accurate in separating the categories of the problem than the rest, and this is reflected in both precision and recall. The key element of the GenClass method is that it can generate classification rules programmatically and detect the hidden associations that exist between the problem features and the desired classes. However, Genetic Programming-based techniques require significantly higher execution time compared to other machine learning techniques. This is most evident in the feature construction method where at each generation of the genetic algorithm, a set of learning models is required to be trained to evaluate the generated artificial features. [ABSTRACT FROM AUTHOR]

Missing rainfall data has been a prevalent issue and primarily interested in hydrology and meteorology. This research aimed to examine the capability of machine learning (ML) and spatial interpolation (SI) methods to estimate missing monthly rainfall data. Six ML algorithms (i.e. multiple linear regression (MLR), M5 model tree (M5), random forest (RF), support vector regression (SVR), multilayer perceptron (MLP), genetic programming (GP)) and four SI methods (i.e. arithmetic average (AA), inverse distance weighting (IDW), correlation coefficient weighted (CCW), normal ratio (NR)) were investigated and compared in their performance. The twelve rainfall stations, located in the Thale Sap Songkhla river basin and nearby basins, were considered as a study case. Tuning hyper-parameters for each ML method was conducted to get the most suitable model for the data sets considered. Three performance criteria matrices (i.e. NSE, OI, and r) were chosen, and the sum of those three performance criteria matrices was introduced for methods' performance comparison. The experimental results pointed out that selecting neighbouring stations were essential when applying SI methods, but not for the ML method. The overall performance showed ML better imputed missing monthly rainfall than SI due to overcoming spatial constraints. GP provided the highest performance by giving NSE = 0.825, OI = 0.877, and r = 0.909 for the training stage. Those values for the testing stage were 0.796, 0.852, and 0.902, respectively. It was followed by SVR-rbf, SVR-poly, and RF. NR provided the best performance among four SI methods, followed by CCW, AA, and IDW. When applying SI methods, it should contemplate a correlation between the target and neighbouring stations greater than 0.80. [ABSTRACT FROM AUTHOR]

Dispatching rules are one of the most widely applied methods for solving Dynamic Job Shop Scheduling problems (DJSSP) in real-world manufacturing systems. Hence, the automated design of effective rules has been an important subject in the scheduling literature for the past several years. High computational requirements and difficulty in interpreting generated rules are limitations of literature methods. Also, feature selection approaches in the field of automated design of scheduling policies have been developed for the tree-based GP approach only. Therefore, the aim of this study is to propose a feature selection approach for the Gene Expression Programming (GEP) algorithm to evolve high-quality rules in simple structures with an affordable computational budget. This integration speeds up the search process by restricting the GP search space using the linear representation of the GEP algorithm and creates concise rules with only meaningful features using the feature selection approach. The proposed algorithm is compared with five algorithms and 30 rules from the literature under different processing conditions. Three performance measures are considered including total weighted tardiness, mean tardiness, and mean flow time. The results show that the proposed algorithm can generate smaller rules with high interpretability in a much shorter training time. [ABSTRACT FROM AUTHOR]

We propose a framework designed to tackle a multi-objective optimization challenge related to the placement of applications in fog computing, employing a deep reinforcement learning (DRL) approach. Unlike other optimization techniques, such as integer linear programming or genetic algorithms, DRL models are applied in real time to solve similar problem situations after training. Our model comprises a learning process featuring a graph neural network and two actor-critics, providing a holistic perspective on the priorities concerning interconnected services that constitute an application. The learning model incorporates the relationships between services as a crucial factor in placement decisions: Services with higher dependencies take precedence in location selection. Our experimental investigation involves illustrative cases where we compare our results with baseline strategies and genetic algorithms. We observed a comparable Pareto set with negligible execution times, measured in the order of milliseconds, in contrast to the hours required by alternative approaches. [ABSTRACT FROM AUTHOR]

Reducing the cost of operating and maintaining wind farms is essential for the economic viability of this renewable energy source. This study applies hyper-heuristics to design a maintenance policy that prescribes the best maintenance action in every possible situation. Genetic programming is used to construct a priority function that determines what maintenance activities to conduct and the sequence of maintenance activities if there are not enough resources to do all of them simultaneously. The priority function may take into account the health condition of the target turbine and its components, the characteristics of the corresponding maintenance work, the workload of the maintenance crew, the working condition of the whole wind farm and the possibilities provided by opportunistic maintenance. Empirical results using a simulation model of the wind farm demonstrate that the proposed model can construct maintenance policies that perform well both in training and test scenarios, which shows the practicability of the approach. [ABSTRACT FROM AUTHOR]

In recent years, the growing interest in environmental sustainability has led to Electric Vehicle Routing Problems (EVRPs) attracting more and more attention. EVRPs involve the use of electric vehicles, which have additional constraints, such as range and recharging time, compared to conventional Vehicle Routing Problems (VRPs). The complexity and dynamic nature of solving VRPs often lead to the introduction of Routing Policies (RPs), simple heuristics that incrementally build routes. However, manually designing efficient RPs proves to be a challenging and time-consuming task. Therefore, there is a pressing need to explore the application of hyper-heuristics, in particular Genetic Programming (GP), to automatically generate new RPs. Since this method has not yet been investigated in the literature in the context of EVRPs, this study explores the applicability of GP to automatically generate new RPs for EVRP. To this end, three RP variants (serial, semiparallel, and parallel) are introduced in this study, along with a set of domain-specific terminal nodes to optimise three criteria: the number of vehicles, energy consumption, and total tardiness. The experimental analysis shows that the serial variant performs best in terms of energy consumption and number of vehicles, while the parallel variant is most effective in minimising the total tardiness. A comprehensive analysis of the proposed method is conducted to determine its convergence properties and the impact of the proposed terminal nodes on performance and to describe several generated RPs. The results show that the automatically generated RPs perform commendably compared to traditional methods such as metaheuristics and exact methods, which usually require significantly more runtime. More specifically, depending on the scenario in which they are used, the generated RPs achieve results that are about 20%-37% worse compared to the best known results for the number of vehicles in almost negligible time, in just some milliseconds. [ABSTRACT FROM AUTHOR]

This paper introduces a cooperative approach of a swarm intelligence algorithm and a linear programming solver to solve the capacitated facility location problem (CFLP). Given a set of potential locations to open facilities, the aim in CFLP is to find the minimum cost, which is the sum of facility opening costs and transportation costs. The developed solution strategy decomposes CFLP into two sub-problems. The former sub-problem has a binary domain. Although most of the swarm intelligence algorithms employ additional procedures such as sigmoid function to deal with binary domains, the proposed algorithm does not require for such methods. An adaptive mutation operator enhances this algorithm. The aim of the latter sub-problem is to generate a policy that optimally assigns customers to the opened facilities. In this regard, the generated binary vectors by the proposed algorithm are passed to a solver to optimise the generated linear model. Commonly used instances available in the literature are solved by the proposed strategy. Comprehensive experimental study includes comparisons with the sate-of-the-art. According to the statistically verified results, the proposed strategy is found as promising in solving CFLP. [ABSTRACT FROM AUTHOR]

Abstract Bentonite plastic concrete (BPC) is extensively used in the construction of water-tight structures like cut-off walls in dams, etc., because it offers high plasticity, improved workability, and homogeneity. Also, bentonite is added to concrete mixes for the adsorption of toxic metals. The modified design of BPC, as compared to normal concrete, requires a reliable tool to predict its strength. Thus, this study presents a novel attempt at the application of two innovative evolutionary techniques known as multi-expression programming (MEP) and gene expression programming (GEP) and a boosting-based algorithm known as AdaBoost to predict the 28-day compressive strength ( ) of BPC based on its mixture composition. The MEP and GEP algorithms expressed their outputs in the form of an empirical equation, while AdaBoost failed to do so. The algorithms were trained using a dataset of 246 points gathered from published literature having six important input factors for predicting. The developed models were subject to error evaluation, and the results revealed that all algorithms satisfied the suggested criteria and had a correlation coefficient (R) greater than 0.9 for both the training and testing phases. However, AdaBoost surpassed both MEP and GEP in terms of accuracy and demonstrated a lower testing RMSE of 1.66 compared to 2.02 for MEP and 2.38 for GEP. Similarly, the objective function value for AdaBoost was 0.10 compared to 0.176 for GEP and 0.16 for MEP, which indicated the overall good performance of AdaBoost compared to the two evolutionary techniques. Also, Shapley additive analysis was done on the AdaBoost model to gain further insights into the prediction process, which revealed that cement, coarse aggregate, and fine aggregate are the most important factors in predicting the strength of BPC. Moreover, an interactive graphical user interface (GUI) has been developed to be practically utilized in the civil engineering industry for prediction of BPC strength.

Abstract Accurately predicting the Modulus of Resilience (MR) of subgrade soils, which exhibit non-linear stress–strain behaviors, is crucial for effective soil assessment. Traditional laboratory techniques for determining MR are often costly and time-consuming. This study explores the efficacy of Genetic Programming (GEP), Multi-Expression Programming (MEP), and Artificial Neural Networks (ANN) in forecasting MR using 2813 data records while considering six key parameters. Several Statistical assessments were utilized to evaluate model accuracy. The results indicate that the GEP model consistently outperforms MEP and ANN models, demonstrating the lowest error metrics and highest correlation indices (R2). During training, the GEP model achieved an R2 value of 0.996, surpassing the MEP (R2 = 0.97) and ANN (R2 = 0.95) models. Sensitivity and SHAP (SHapley Additive exPlanations) analysis were also performed to gain insights into input parameter significance. Sensitivity analysis revealed that confining stress (21.6%) and dry density (26.89%) are the most influential parameters in predicting MR. SHAP analysis corroborated these findings, highlighting the critical impact of these parameters on model predictions. This study underscores the reliability of GEP as a robust tool for precise MR prediction in subgrade soil applications, providing valuable insights into model performance and parameter significance across various machine-learning (ML) approaches.

Fault classification in industrial processes is of paramount importance, as it allows the implementation of preventive and corrective measures before catastrophic failures occur, which can result in significant repair costs and production loss, for example. Therefore, the purpose of this study was to develop a classification model by merging the concepts of Decision Trees with Genetic Programming. To accomplish this, the proposed model randomly generates a set of decision trees using the adapted Tennessee Eastman dataset. The generation of these trees does not rely on classical construction logic; instead, they employ an approach where the structure and characteristics of the trees are randomly determined and adjusted throughout the evolutionary process. This approach enables a broader exploration of the search space and may lead to diverse solutions. The results obtained were moderate, largely due to the high number of target classes for classification (21 classes), resulting in the creation of complex trees. The average accuracy on the test data was 0.75, indicating the need to implement new alternatives and enhancements in the algorithm to improve the results.

Abstract Background North African human populations present a complex demographic scenario due to the presence of an autochthonous genetic component and population substructure, plus extensive gene flow from the Middle East, Europe, and sub-Saharan Africa. Results We conducted a comprehensive analysis of 364 genomes to construct detailed demographic models for the North African region, encompassing its two primary ethnic groups, the Arab and Amazigh populations. This was achieved through an Approximate Bayesian Computation with Deep Learning (ABC-DL) framework and a novel algorithm called Genetic Programming for Population Genetics (GP4PG). This innovative approach enabled us to effectively model intricate demographic scenarios, utilizing a subset of 16 whole genomes at > 30X coverage. The demographic model suggested by GP4PG exhibited a closer alignment with the observed data compared to the ABC-DL model. Both point to a back-to-Africa origin of North African individuals and a close relationship with Eurasian populations. Results support different origins for Amazigh and Arab populations, with Amazigh populations originating back in Epipaleolithic times, while GP4PG supports Arabization as the main source of Middle Eastern ancestry. The GP4PG model includes population substructure in surrounding populations (sub-Saharan Africa and Middle East) with continuous decaying gene flow after population split. Contrary to ABC-DL, the best GP4PG model does not require pulses of admixture from surrounding populations into North Africa pointing to soft splits as drivers of divergence in North Africa. Conclusions We have built a demographic model on North Africa that points to a back-to-Africa expansion and a differential origin between Arab and Amazigh populations.

Abstract The utilization of Self-compacting Concrete (SCC) has escalated worldwide due to its superior properties in comparison to normal concrete such as compaction without vibration, increased flowability and segregation resistance. Various other desirable properties like ductile behaviour, increased strain capacity and tensile strength etc. can be imparted to SCC by incorporation of fibres. Thus, this study presents a novel approach to predict 28-day compressive strength (C–S) of FR-SCC using Gene Expression Programming (GEP) and Multi Expression Programming (MEP) for fostering its widespread use in the industry. For this purpose, a dataset had been compiled from internationally published literature having six input parameters including water-to-cement ratio, silica fume, fine aggregate, coarse aggregate, fibre, and superplasticizer. The predictive abilities of developed algorithms were assessed using error metrices like mean absolute error (MAE), a20-index, and objective function (OF) etc. The comparison of MEP and GEP models indicated that GEP gave a simple equation having lesser errors than MEP. The OF value of GEP was 0.029 compared to 0.031 of MEP. Thus, sensitivity analysis was performed on GEP model. The models were also checked using some external validation checks which also verified that MEP and GEP equations can be used to forecast the strength of FR-SCC for practical uses.

Recognizing the differential impacts of climate change across geographical scales, this study emphasizes the importance of statistical downscaling. Using Gene Expression Programming (GEP) and Linear Genetic Programming (LGP), statistical downscaling transforms broad climate trends into region-specific insights. This allowed for detailed analyses of anticipated changes in sediment yield and discharge within a Euphrates River sub-basin in Türkiye using large-scale variables from the CanESM2 model. The dataset is divided into calibration (1970–1995) and validation (1996–2005) periods. To assess the models’ accuracy, statistical measures such as RMSE, MAE, NSE, and R were used. The analysis revealed that LGP outperformed GEP in both discharge and sediment yield during validation, with RMSE = 51.79 m3/s and 4,325.66 tons/day, MAE = 27.14 m3/s and 1,593.34 tons/day, NSE = 0.684 and 0.627, and R = 0.841 and 0.788, respectively. However, when simulating future periods based on the observed period (2006–2020), the GEP model was superior to LGP under RCP2.6, RCP4.5, and RCP 8.5 scenarios from CanESM2. In 2021–2100, models suggest a moderate decrease in discharge and sediment yield, indicating potential shifts in the basin's hydrodynamics. These changes could disrupt hydropower generation, challenge water management practices, and alter riverine ecosystems. The results necessitate a thorough assessment of potential ecological consequences. HIGHLIGHTS Climate change effects on streamflow and sediment yield have been explored.; Genetic programming such as GEP and LGP have been employed in statistical downscaling technique.; Future projection of streamflow and sediment yield have been done under different emission scenarios.;

Abstract Bitumen, aggregate, and air void (VA) are the three primary ingredients of asphalt concrete. VA changes over time as a function of four factors: traffic loads and repetitions, environmental regimes, compaction, and asphalt mix composition. Due to the high as-constructed VA content of the material, it is expected that VA will reduce over time, causing rutting during initial traffic periods. Eventually, the material will undergo shear flow when it reaches its densest state with optimum aggregate interlock or refusal VA content. Therefore, to ensure the quality of construction, VA in asphalt mixture need to be modeled throughout the service life. This study aims to implement a hybrid evolutionary polynomial regression (EPR) combined with a teaching–learning based optimization (TLBO) algorithm and multi-gene genetic programming (MGGP) to predict the VA percentage of asphalt mixture during the service life. For this purpose, 324 data records of VA were collected from the literature. The variables selected as inputs were original as-constructed VA, $${VA}_{orig}$$ VA orig (%); mean annual air temperature, $$MAAT$$ MAAT (°F); original viscosity at 77 °F, $${\eta }_{orig,77}$$ η o r i g , 77 (Mega-Poises); and $$time$$ time (months). EPR-TLBO was found to be superior to MGGP and existing empirical models due to the interquartile ranges of absolute error boxes equal to 0.67%. EPR-TLBO had an R2 value of more than 0.90 in both the training and testing phases, and only less than 20% of the records were predicted utilizing this model with more than 20% deviation from the observed values. As determined by the sensitivity analysis, $${\eta }_{orig,77}$$ η o r i g , 77 is the most significant of the four input variables, while time is the least one. A parametric study showed that regardless of $$MAAT$$ MAAT , $${\eta }_{orig,77},$$ η o r i g , 77 , of 0.3 Mega-Poises, and $${VA}_{orig}$$ VA orig above 6% can be ideal for improving the pavement service life. It was also witnessed that with an increase of $$MAAT$$ MAAT from 37 to 75 °F, the serviceability of asphalt concrete takes 15 months less on average.

Abstract In the paper, a new evolutionary technique called Linear Matrix Genetic Programming (LMGP) is proposed. It is a matrix extension of Linear Genetic Programming and its application is data-driven black-box control-oriented modeling in conditions of limited access to training data. In LMGP, the model is in the form of an evolutionarily-shaped program which is a sequence of matrix operations. Since the program has a hidden state, running it for a sequence of input data has a similar effect to using well-known recurrent neural networks such as Long Short-Term Memory (LSTM) or Gated Recurrent Unit (GRU). To verify the effectiveness of the LMGP, it was compared with different types of neural networks. The task of all the compared techniques was to reproduce the behavior of a nonlinear model of an underwater vehicle. The results of the comparative tests are reported in the paper and they show that the LMGP can quickly find an effective and very simple solution to the given problem. Moreover, a detailed comparison of models, generated by LMGP and LSTM/GRU, revealed that the former are up to four times more accurate than the latter in reproducing vehicle behavior.

Nitrate is the primary water-soluble macronutrient essential for plant growth that is converted from excess fish feeds, fish effluents, and degrading biomaterials on the aquaponic pond floor, and when aquacultural malpractices occur, large amounts of it retain in the water system causing increase rate in eutrophication and toxifies fish and aquaculture plants. Recent nitrate sensor prototypes still require performing the additional steps of water sample deionization and dilution and were constructed with expensive materials. In response to the challenge of sensor enhancement and aquaponic water quality monitoring, this study developed sensitive, repeatable, and reproducible screen-printed graphite electrodes on polyvinyl chloride and parchment paper substrates with silver as electrode material and 60:40 graphite powder:nail polish formulated conductive ink for electrical traces, integrated with 9-gene genetic expression model as a function of peak anodic current and electrochemical test time for nitrate concentration prediction that is embedded into low-power Arduino ESP32 for in situ nitrate sensing in aquaponic pond water. Five SPE electrical traces were designed on the two types of substrates. Scanning electron microscopy with energy dispersive X-ray confirmed the electrode surface morphology. Electrochemical cyclic voltammetry using 10 to 100 mg/L KNO3 and water from three-depth regions of the actual pond established the electrochemical test time (10.5 s) and electrode potential (0.135 V) protocol necessary to produce peak current that corresponds to the strength of nitrate ions during redox. The findings from in situ testing revealed that the proposed sensors have strong linear predictions (R2 = 0.968 MSE = 1.659 for nSPEv and R2 = 0.966 MSE = 4.697 for nSPEp) in the range of 10 to 100 mg/L and best detection limit of 3.15 μg/L, which are comparable to other sensors of more complex construction. The developed three-electrode electrochemical nitrate sensor confirms that it is reliable for both biosensing in controlled solutions and in situ aquaponic pond water systems.

Assessing ecosystem health on a large scale is crucial for a wide range of management and regulatory decisions. Technologies such as hyperspectral imaging allow noninvasive and rapid estimation of key attributes based on observed reflectance. However, these images are high-dimensional and real-world applications require models based on fewer wavelengths. This paper proposes a new wavelength selection and feature extraction method for hyperspectral image analysis based on genetic programming to automatically select key wavelength regions and informative image features. A dataset of hyperspectral images of sediment in the field was collected and paired with ground-truth measurements of the sediment porosity and organic matter content. Two new program structures were proposed to construct feature extraction trees from either the mean reflectance spectra (spectra-based) or full hyperspectral images (image-based). SVR models were constructed to predict attributes based on the extracted features. Various regression models were used to predict the porosity and organic matter content. Full-wavelength models were constructed to reliably predict the organic matter content. The proposed spectra-based genetic programming solutions show competitive results compared to common wavelength selection methods, such as SPA, CARS, and RC. Finally, the best-evolved solution was applied to predict sediment organic matter content across all collected images. [ABSTRACT FROM AUTHOR]

Laser powder bed fusion (LPBF) is a metal additive manufacturing (AM) process where a part is created layer-by-layer through the melting and solidification of powder layers. The main challenge in LPBF is quality assurance. An estimation of powder rheology/flowability, as a quality-impacting factor, is therefore necessary prior to printing. To predict powder flowability based on morphological features without cost- and time-inefficient experiments, this study explores using machine learning algorithms (ML). It focuses on a sample set of standard and off-size Ti6Al4V (Ti64) powders, aiming to (1) assess ML regressors and equation fitting for flowability prediction, using the small original dataset and (2) overcome data limitations' impact on regression through data augmentation techniques. Results showed conventional ML regressors like support vector regression (SVR) sufficed for dynamic flowability prediction. Still, more complex targets such as shear flowability and hall flowability required innovative approaches like genetic programming (equation fitting) and synthetic minority over-sampling (SMOTE), respectively, as proposed in this work. [ABSTRACT FROM AUTHOR]

Artificial Intelligence (AI) methods, including Artificial Neural Networks (ANNs), Adaptive Neuro-Fuzzy Inference Systems (ANFISs), Support Vector Machines (SVMs), Deep Learning (DL), Genetic Programming (GP) and Hybrid Algorithms, have proven to be important tools for accurate groundwater level (GWL) modelling. Through an analysis of the results obtained in numerous articles published in high-impact journals during 2001–2023, this comprehensive review examines each method's capabilities, their combinations, and critical considerations about selecting appropriate input parameters, using optimisation algorithms, and considering the natural physical conditions of the territories under investigation to improve the models' accuracy. For example, ANN takes advantage of its ability to recognise complex patterns and non-linear relationships between input and output variables. In addition, ANFIS shows potential in processing diverse environmental data and offers higher accuracy than alternative methods such as ANN, SVM, and GP. SVM excels at efficiently modelling complex relationships and heterogeneous data. Meanwhile, DL methods, such as Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNNs), are crucial in improving prediction accuracy at different temporal and spatial scales. GP methods have also shown promise in modelling complex and nonlinear relationships in groundwater data, providing more accurate and reliable predictions when combined with optimisation techniques and uncertainty analysis. Therefore, integrating these methods and optimisation techniques (Hybrid Algorithms), tailored to specific hydrological and hydrogeological conditions, can significantly increase the predictive capability of GWL models and improve the planning and management of water resources. These findings emphasise the importance of thoroughly understanding (a priori) the functionalities and capabilities of each potentially beneficial AI-based methodology, along with the knowledge of the physical characteristics of the territory under investigation, to optimise GWL predictive models. [ABSTRACT FROM AUTHOR]

Machine learning (ML) is commonly used to develop survival-predictive radiomic models in non-small cell lung cancer (NSCLC) patients, which helps assist treatment decision making. Radiomic features derived from computer tomography (CT) lung images aim to capture quantitative tumor characteristics. However, these features are determined by humans, which poses a risk of including irrelevant or redundant variables, thus reducing the model's generalization. To address this issue, we propose using genetic programming (GP) to automatically construct new features with higher discriminant power than the original radiomic features. To achieve this goal, we introduce a fitness function that measures the classification performance ratio of output to input. The constructed features are then input for various classifiers to predict the two-year survival of NSCLC patients from two public CT datasets. Our approach is compared against two popular feature selection methods in radiomics to choose relevant radiomic features, and two GP-based feature construction methods whose fitness functions are based on measuring the constructed features' quality. The experimental results show that survival prediction models trained on GP-based constructed features outperform feature selection methods. Also, maximizing the classification performance gain output-to-input ratio produces features with higher discriminative power than only maximizing the classification accuracy from constructed features. Furthermore, a survival analysis demonstrated statistically significant differences between survival and non-survival groups in the Kaplan–Meier curves. Therefore, the proposed approach can be used as a complementary method for oncologists in determining the clinical management of NSCLC patients. [ABSTRACT FROM AUTHOR]

We present a machine learning framework capable of consistently inferring mathematical expressions of hyperelastic energy functionals for incompressible materials from sparse experimental data and physical laws. To achieve this goal, we propose a polyconvex neural additive model (PNAM) that enables us to express the hyperelastic model in a learnable feature space while enforcing polyconvexity. An upshot of this feature space obtained via the PNAM is that (1) it is spanned by a set of univariate basis functions that can be re‐parametrized with a more complex mathematical form, and (2) the resultant elasticity model is guaranteed to fulfill the polyconvexity, which ensures that the acoustic tensor remains elliptic for any deformation. To further improve the interpretability, we use genetic programming to convert each univariate basis into a compact mathematical expression. The resultant multi‐variable mathematical models obtained from this proposed framework are not only more interpretable but are also proven to fulfill physical laws. By controlling the compactness of the learned symbolic form, the machine learning‐generated mathematical model also requires fewer arithmetic operations than its deep neural network counterparts during deployment. This latter attribute is crucial for scaling large‐scale simulations where the constitutive responses of every integration point must be updated within each incremental time step. We compare our proposed model discovery framework against other state‐of‐the‐art alternatives to assess the robustness and efficiency of the training algorithms and examine the trade‐off between interpretability, accuracy, and precision of the learned symbolic hyperelastic models obtained from different approaches. Our numerical results suggest that our approach extrapolates well outside the training data regime due to the precise incorporation of physics‐based knowledge. [ABSTRACT FROM AUTHOR]