Search

The layout, scale and spatial form of urban employment centers are important guidelines for the rational layout of public service facilities such as urban transportation, medical care, and education. In this paper, we use Internet cell phone positioning data to identify the workplace and residence of users in the Beijing city area and obtain commuting data of the employed to measure the employment center system in Beijing. Firstly, the employment density distribution is generated using the data of the working places of the employed persons, and the employment centers are identified based on the employment density of Beijing. Then, we use the business registration data of employment centers to measure the industrial diversity within the employment centers by using the ecological Shannon Wiener Diversity Index, and combine the commuting links between employment centers and places of residence to measure the energy level of each employment center, analyze the hinterland and sphere of influence of each center, and finally using the industrial diversity index of employment centers and the average commuting time of employed persons, combined with the K-Means clustering algorithm, to classify the employment centers in Beijing. The employment center identification and classification method based on big data constructed in this study can help solve the limitations of the previous employment center system research in terms of center identification and commuting linkage measurement due to large spatial units and lack of commuting data to a certain extent. The study can provide reference for the regular understanding and technical analysis of employment centers and provide help for the employment multi-center system in Beijing in terms of quantifying the employment spatial structure, guiding the construction of multi-center system, and adjusting the land use rules. [ABSTRACT FROM AUTHOR]

Big data analytics is an emerging topic in academic and industrial engineering fields, where the large-scale data issue is the most attractive challenge. It is crucial to design an effective large-scale data processing model to handle big data. In this paper, we aim to improve the accuracy of the classification task and reduce the execution time for large-scale data within a small cluster. In order to overcome these challenges, this paper presents a novel ensemble-based paradigm that consists of the procedure of splitting large-scale data files and developing ensemble models. Two different splitting methods are first developed to partition large-scale data into small data blocks without overlapping. Then we propose two ensemble-based methods with high accuracy and less execution time: bagging-based and boosting-based methods. Finally, the proposed paradigm can be implemented by four predictive models, which are combinations of two splitting methods and two ensemble-based methods. A series of persuasive experiments was conducted to evaluate the effectiveness of the proposed paradigm with four different combinations. Overall, the proposed paradigm with boosting-based is the best in terms of the accuracy metric compared with existing methods. In addition, boosting-based methods achieve 91.6% accuracy compared with 52% accuracy of base line model for a big data file with 10 million samples. However, the paradigm with bagging-based takes the least execution time to yield results. This paper also reveals the effectiveness of the computing Spark cluster for large-scale data and points out the weakness of RDD (Resilient Distributed dataset). [ABSTRACT FROM AUTHOR]

Feature Selection (FS), a pre-processing step used in the majority of big data processing applications, aims to eliminate irrelevant and redundant features from the data. Its purpose is to select a final set of data characteristics that best represent the data as a whole. To achieve this, it explores every potential solution in order to identify the optimal subset. Meta-heuristic algorithms have been found to be particularly effective in solving FS problems, especially for high-dimensional datasets. This work adopts a recently developed meta-heuristic called the White Shark Optimizer (WSO) due to its simplicity and low computational overhead. However, WSO faces challenges in effectively balancing exploration and exploitation, particularly in complex multi-peak search problems. It tends to converge prematurely and get stuck in local optima, which can lead to poor search performance when dealing with FS problems. To overcome these issues, this paper presents three enhanced binary variants of WSO for well-known FS problems. These variants are as follows: (1) Binary Distribution-based WSO (BDWSO), where the algorithm refines the positions of white sharks by considering both the average and standard deviation of the current shark, the local best shark, and the global best shark. This strategy is designed to alleviate issues of premature convergence and stagnation during iterations , (2) Binary Sine Cosine WSO (BSCWSO), which uses sine and cosine adaptive functions for the social and cognitive components of the position update rule, and (3) Binary Hybrid Sine Cosine WSO (BHSCWSO), which employs sine and cosine acceleration factors to regulate local search and achieve convergence to the global optimal solution. Additionally, the population was initialized using the Opposition-Based Learning (OBL) mechanism, and the sine map was used to modify the inertia weight of WSO. These revised variants of WSO were established to have a better harmony between the exploration and exploitation facets. The proposed methods were extensively compared to the fundamental binary WSO and other well-known algorithms in the field. The experimental findings and comparisons demonstrate that the proposed methods outperform the conventional and most evaluated similar algorithms in terms of robustness and solution quality. In terms of classification accuracy, number of selected features, specificity, sensitivity, and fitness values, the proposed BHSCWSO optimizer performed better than all other proposed peer optimizers, including BSWO, BDWSO, and BSCWSO, in 11, 8, 13, 18, and 10 datasets, respectively. The proposed BHSCWSO optimizer showed performance levels of more than 90% in terms of accuracy, sensitivity, and specificity metric measures on 15, 14, and 11 of the 24 datasets deemed. [ABSTRACT FROM AUTHOR]

This paper identifies trends in the application of big data in the transport sector and categorizes research work across scientific subfields. The systematic analysis considered literature published between 2012 and 2022. A total of 2671 studies were evaluated from a dataset of 3532 collected papers, and bibliometric techniques were applied to capture the evolution of research interest over the years and identify the most influential studies. The proposed unsupervised classification model defined categories and classified the relevant articles based on their particular scientific interest using representative keywords from the title, abstract, and keywords (referred to as top words). The model's performance was verified with an accuracy of 91% using Naïve Bayesian and Convolutional Neural Networks approach. The analysis identified eight research topics, with urban transport planning and smart city applications being the dominant categories. This paper contributes to the literature by proposing a methodology for literature analysis, identifying emerging scientific areas, and highlighting potential directions for future research. [ABSTRACT FROM AUTHOR]

Data classification is a crucial aspect of knowledge discovery using machinelearning algorithm for supervised learning approach where the goal is to predict the categorical labels of new instances based on past observations. This research presents an innovative classification technique that utilizes Rough Set Attribute Reduction. The proposed method introduces the Re-heat Simulated Annealing (Re-heat SA) algorithm as a meta-heuristic approach. Rough set theory, a mathematical tool dealing with uncertainty and fuzziness in data, is employed to uncover hidden patterns in big data through feature selection. This paper introduces a novel meta-heuristic classification approach that utilizes rough set attribute reduction to achieve optimal accuracy. Re-heat SA effectively optimizes the problem by controlling the dependency degree to identify the minimal reducts required for classification prediction using the Rosetta software. Experimental results demonstrate that Re-heat SA outperforms comparable classification algorithms in discovering classification rules. The results reveal that three datasets achieved 100% accuracy, four datasets achieved accuracy rates ranging from 60% to 99%, and six datasets achieved accuracy rates ranging from 30% to 59%. Additionally, this paper discusses the need for standardization concerning the machine learning pipeline processes as big data and its handling grows exponentially. [ABSTRACT FROM AUTHOR]

Differences in data size per class, also known as imbalanced data distribution, have become a common problem affecting data quality. Big Data scenarios pose a new challenge to traditional imbalanced classification algorithms, since they are not prepared to work with such amount of data. Split data strategies and lack of data in the minority class due to the use of MapReduce paradigm have posed new challenges for tackling the imbalance between classes in Big Data scenarios. Ensembles have been shown to be able to successfully address imbalanced data problems. Smart Data refers to data of enough quality to achieve high-performance models. The combination of ensembles and Smart Data, achieved through Big Data preprocessing, should be a great synergy. In this paper, we propose a novel Smart Data driven Decision Trees Ensemble methodology for addressing the imbalanced classification problem in Big Data domains, namely SD_DeTE methodology. This methodology is based on the learning of different decision trees using distributed quality data for the ensemble process. This quality data is achieved by fusing random discretization, principal components analysis, and clustering-based random oversampling for obtaining different Smart Data versions of the original data. Experiments carried out in 21 binary adapted datasets have shown that our methodology outperforms random forest. [ABSTRACT FROM AUTHOR]

Diabetes is regarded as one of the deadliest chronic illnesses that increases blood sugar. But there is no reliable method for predicting diabetic severity that shows how the disease will affect various body organs in the future. Therefore, this paper introduced Optimized Dual Directional Temporal convolution and Attention based Density Clustering (ODDTADC) method for predicting and classifying risk level in diabetic patients. In the diabetic prediction stage, the prediction is done by using an Integrated Dual Directional Temporal Convolution and an Enriched Remora Optimization Algorithm. Here, dual directional temporal convolution is used to extract temporal features by integrating dilated convolution and casual convolution in the feature extraction layer. Then, the attention module is used instead of max-pooling to emphasize the various features' importance in the feature aggregation layer. The Enriched Remora Optimization Algorithm is used to find optimal hyper parameters for Integrated Dual Directional Temporal Convolution. In the classification of stages based on risk level, the values from stage-I are fed into the Attention based Density Spatial Clustering of Applications with Noise, which allocate various weights based on their density values in the Core Points. Based on the results, the Nested Long Short-Term Memory is utilized to classify the risk levels of diabetic patients over a period of two or three years. Experimental evaluations were performed on five datasets, including PIMA Indian Diabetics Database, UCI Machine Learning Repository Diabetics Dataset, Heart Diseases Dataset, Chronic Disease Dataset and Diabetic Retinopathy Debrecen Dataset. The proposed ODDTADC method demonstrates superior performance compared to existing methods, achieving remarkable results in accuracy (98.21%), recall (94.46%), kappa coefficient (98.95%), precision (98.74%), F1-score (99.01%) and Matthew’s correlation coefficient (MCC) (0.87%). [ABSTRACT FROM AUTHOR]

A great deal of study has been done recently on support vector machines (SVMs) and how they are used in various scientific domains. The statistical learning theory's mathematical foundation allows SVM to offer logical solutions to issues. A portion of the training input serves as the SVM's solution. SVM is frequently employed in applications involving feature reduction. regression. and novelty detection. Additionally, studies conducted in certain fields where SVM performs badly have prompted the creation of alternative SVM applications. including big data sets, multiclassification SVM. and imbalanced data sets SVM. Furthermore, SVM is used with other more sophisticated techniques. such as evolve algorithms. to enhance classification performance and optimize parameters. SVM algorithms are now widely used in research and applications across a number of scientific and engineering domains. The SVM might be one of the primary options for big data compatibility and big data classification because of its advantages. Data preparation methods must be created in order to mask data into the appropriate format for learning in order to do this. This paper provides a quick overview of SVM. list several applications, draw attention to current issues and patterns, and point out sonic of SVM's drawbacks in this essay. This document can be used to categorize vast volumes of data and recommend research areas for further investigation. [ABSTRACT FROM AUTHOR]

Wavelet neural networks (WNN) have been applied in many fields to solve regression as well as classification problems. After the advent of big data, as data gets generated at a brisk pace, it is imperative to analyze it as soon as it is generated owing to the fact that the nature of the data may change dramatically in short time intervals. This is necessitated by the fact that big data is all pervasive and throws computational challenges for data scientists. Therefore, in this paper, we built an efficient Scalable, Parallelized Wavelet Neural Network (SPWNN) which employs the parallel stochastic gradient algorithm (SGD) algorithm. SPWNN is designed and developed under both static and streaming environments in the horizontal parallelization framework. SPWNN is implemented by using Morlet and Gaussian functions as activation functions. This study is conducted on big datasets like gas sensor data which has more than 4 million samples and medical research data which has more than 10,000 features, which are high dimensional in nature. The experimental analysis indicates that in the static environment, SPWNN with Morlet activation function outperformed SPWNN with Gaussian on the classification datasets. However, in the case of regression, there is no clear trend was observed. In contrast, in the streaming environment i.e., Gaussian outperformed Morlet on the classification and Morlet outperformed Gaussian on the regression datasets. Overall, the proposed SPWNN architecture achieved a speedup of 1.22 - 1.78. [ABSTRACT FROM AUTHOR]

One of the greatest challenges of our time is monitoring the rapid environmental changes taking place worldwide at both local and global scales. This requires easy-to-use and ready-to-implement tools and services to monitor and quantify aspects of bio- and geodiversity change and the impact of land use intensification using freely available and global remotely sensed data, and to derive remotely sensed indicators. Currently, there are no services for quantifying both raster- and vector-based indicators in a "compact tool". Therefore, the main innovation of ESIS/Imalys is having a remote sensing (RS) tool that allows for RS data processing, data management, and continuous and discrete quantification and derivation of RS indicators in one tool. With the ESIS/Imalys project (Ecosystem Integrity Remote Sensing—Modelling and Service Tool), we try to present environmental indicators on a clearly defined and reproducible basis. The Imalys software library generates the RS indicators and remote sensing products defined for ESIS. This paper provides an overview of the functionality of the Imalys software library. An overview of the technical background of the implementation of the Imalys library, data formats and the user interfaces is given. Examples of RS-based indicators derived using the Imalys tool at pixel level and at zone level (vector level) are presented. Furthermore, the advantages and disadvantages of the Imalys tool are discussed in detail in order to better assess the value of Imalys for users and developers. The applicability of the indicators will be demonstrated through three ecological applications, namely: (1) monitoring landscape diversity, (2) monitoring landscape structure and landscape fragmentation, and (3) monitoring land use intensity and its impact on ecosystem functions. Despite the integration of large amounts of data, Imalys can run on any PC, as the processing and derivation of indicators has been greatly optimised. The Imalys source code is freely available and is hosted and maintained under an open source license. Complete documentation of all methods, functions and derived indicators can be found in the freely available Imalys manual. The user-friendliness of Imalys, despite the integration of a large amount of RS data, makes it another important tool for ecological research, modelling and application for the monitoring and derivation of ecosystem indicators from local to global scale. [ABSTRACT FROM AUTHOR]

Ephemeral Fast Radio Bursts (FRBs) must be powered by some of the most energetic processes in the Universe. That makes them highly interesting in their own right, and as precise probes for estimating cosmological parameters. This field thus poses a unique challenge: FRBs must be detected promptly and immediately localised and studied based only on that single millisecond-duration flash. The problem is that the burst occurrence is highly unpredictable and that their distance strongly suppresses their brightness. Since the discovery of FRBs in single-dish archival data in 2007, detection software has evolved tremendously. Pipelines now detect bursts in real time within a matter of seconds, operate on interferometers, buffer high-time and frequency resolution data, and issue real-time alerts to other observatories for rapid multi-wavelength follow-up. In this paper, we review the components that comprise a FRB search software pipeline, we discuss the proven techniques that were adopted from pulsar searches, we highlight newer, more efficient techniques for detecting FRBs, and we conclude by discussing the proposed novel future methodologies that may power the search for FRBs in the era of big data astronomy. [ABSTRACT FROM AUTHOR]

The diversification of business forms leads to functional and spatial complexity in cities. The efficient determination of the complexity of an urban system is the basis for the scientific monitoring of the multi-functional aggregation within cities. Previous studies on the urban spatial structure were limited by the difficulty of collecting micro-data and the high time cost, and they focused on the macro-spatial structure, lacking fine-grained investigations of the micro-spatial structure. Additionally, high-resolution remote sensing images, which mainly rely on the textural characteristics of the spectrum of ground objects, cannot detect the social and economic functions of ground objects. Thus, it is difficult to meet the actual needs of urban planning and management. The purpose of this paper is to automatically identify the spatial heterogeneity and temporal variation of urban land use functions in the context of complex urban systems. The TF-IDF (term frequency–inverse document frequency) algorithm, a machine learning classification algorithm, and other methods are applied to identify the urban functions and distribution characteristics of the main urban area based on the POI (point of interest) data and urban form data. The results show the following: (1) From 2012 to 2022, all types of land use in Yiwu city grew at different rates, with logistics and warehousing space growing the fastest, which is in line with Yiwu's goal of building a national logistics center for trade and services. (2) The residential area has a spatial structure with a dense central circle and a scattered periphery extending from northeast to southwest and from east to west. (3) The commercial service sector shows clear spatial differentiation between the core and the periphery. The commercial functional areas of Niansanli, Houzhai, and Chengxi, where the number of commercial POIs is relatively small, are located at the intersection of the administrative subdistricts near the city center, indicating that the commercial economic activities of the downtown subdistrict have a certain spillover effect on adjacent subdistricts. (4) The public facilities of each subdistrict are generally located in the core of each subdistrict, which ensures better convenience and accessibility. (5) Industrial land with a large total area that is scattered and mixed with urban residential land gradually tends to be centralized, forming an industrial belt around the city. This study comprehensively considers the aggregation relationship between urban buildings and land use and improves the accuracy of land identification and functional zoning. [ABSTRACT FROM AUTHOR]

This paper targets the area of optimizing machine learning (ML) training data by constructing compact data. The methods of optimizing ML training have improved and become a part of artificial intelligence (AI) system development. Compact data learning (CDL) is an alternative practical framework to optimize a classification system by reducing the size of the training dataset. CDL originated from compact data design, which provides the best assets without handling complex big data. CDL is a dedicated framework for improving the speed of the machine learning training phase without affecting the accuracy of the system. The performance of an ML-based arrhythmia detection system and its variants with CDL maintained the same statistical accuracy. ML training with CDL could be maximized by applying an 85% reduced input dataset, which indicated that a trained ML system could have the same statistical accuracy by only using 15% of the original training dataset. [ABSTRACT FROM AUTHOR]

Quantification of all types of uncertainty helps to establish reliability in any analysis. This research focuses on uncertainty in two attribute levels of wetland classification and creates visualization tools to guide analysis of spatial uncertainty patterns over several scales. A novel variant of confusion matrix analysis compares the Cowardin and Hydrogeomorphic wetland classification systems, identifying areas and types of misclassification for binary and multivariate categories. The specific focus on uncertainty in the paper refers to categorical consistency, that is, agreement between the two classification systems, rather than comparing observed data to ground truth. Consistency is quantified using confusion matrix analysis. Aggregation across progressive focal windows transforms the confusion matrix into a multiscale data pyramid for quick determination of where attribute uncertainty is highly variant, and at what spatial resolutions classification inconsistencies emerge. The focal pyramids summarize precision, recall, and F1 scores to visualize classification differences across spatial scales. Findings show that the F1 scores appear most informative on agreement about wetlands misclassification at both coarse and fine attribute scales. The pyramid organizes multi-scale uncertainty in a single unified framework and can be "sliced" to view individual focal levels of attribute consistency. Results demonstrate how the confusion matrix can be used to quantify the percentage of a study area in which inconsistencies occur reflecting wetland presence and type. The research provides confusion metrics and display tools to focus attention on specific areas of large data sets where attribute uncertainty patterns may be complex, thus reducing land managers' workloads by highlighting areas of uncertainty where field checking might be appropriate, and improving analytics by providing visualization tools to quickly see where such areas occur. [ABSTRACT FROM AUTHOR]

Notably, real problems are increasingly complex and require sophisticated models and algorithms capable of quickly dealing with large data sets and finding optimal solutions. However, there is no perfect method or algorithm; all of them have some limitations that can be mitigated or eliminated by combining the skills of different methodologies. In this way, it is expected to develop hybrid algorithms that can take advantage of the potential and particularities of each method (optimization and machine learning) to integrate methodologies and make them more efficient. This paper presents an extensive systematic and bibliometric literature review on hybrid methods involving optimization and machine learning techniques for clustering and classification. It aims to identify the potential of methods and algorithms to overcome the difficulties of one or both methodologies when combined. After the description of optimization and machine learning methods, a numerical overview of the works published since 1970 is presented. Moreover, an in-depth state-of-art review over the last three years is presented. Furthermore, a SWOT analysis of the ten most cited algorithms of the collected database is performed, investigating the strengths and weaknesses of the pure algorithms and detaching the opportunities and threats that have been explored with hybrid methods. Thus, with this investigation, it was possible to highlight the most notable works and discoveries involving hybrid methods in terms of clustering and classification and also point out the difficulties of the pure methods and algorithms that can be strengthened through the inspirations of other methodologies; they are hybrid methods. [ABSTRACT FROM AUTHOR]

The current state of economic, social ideas, and the advancement of cutting-edge technology are determined by the primary subjects of the contemporary information era, big data. People are immersed in a world of information, guided by the abundance of data that penetrates every element of their surroundings. Smart gadgets, the IoT, and other technologies are responsible for the data's explosive expansion. Organisations have struggled to store data effectively throughout the past few decades. This disadvantage is related to outdated, expensive, and inadequately large storage technology. In the meanwhile, large data demands innovative storage techniques supported by strong technology. This paper proposes the bigdata clustering and classification model with improved fuzzy-based Deep Architecture under the Map Reduce framework. At first, the pre-processing phase involves data partitioning from the big dataset utilizing an improved C-Means clustering procedure. The pre-processed big data is then handled by the Map Reduce framework, which involves the mapper and reducer phases. In the mapper phase. Data normalization takes place, followed by the feature fusion approach that combines the extracted features like entropy-based features and correlation-based features. In the reduction phase, all the mappers are combined to produce an acceptable feature. Finally, a deep hybrid model, which is the combination of a DCNN and Bi-GRU is used for the classification process. The Improved score level fusion procedure is used in this case to obtain the final classification result. Moreover, the analysis of the proposed work has proved to be efficient in terms of classification accuracy, precision, recall, FNR, FPR, and other performance metrics. [ABSTRACT FROM AUTHOR]

Imbalanced class distributions in machine learning, where the minority class is often under-represented, pose a substantial challenge. Synthetic Minority Over-sampling Technique (SMOTE) has been widely employed to address this issue by generating synthetic minority samples through interpolation. Despite its popularity, SMOTE exhibits certain drawbacks caused by the implementation of random interpolation samples. In this paper, we introduce a new data level technique for oversampling, called Fuzzy C-Means Center-SMOTE (FCM-CSMOTE), which generates synthetic samples in each cluster using its center considered as the memory of the main data components. We demonstrate that the proposed selective strategy has a very low probability to generate noise. The experimental results demonstrate that the proposed method performs better than the state-of-the-art approaches on 21 real unbalanced data sets (regular and large size data set) in terms of several metrics, including Geometric Mean (GM), F-Measure (FM), Area Under the Curve (AUC), and Accuracy. [ABSTRACT FROM AUTHOR]

Hybridization of classifiers can often yield outperformed performance compared to its best individual component and mostly have more generalization ability. The majority of combined classifiers reported in the literature benefit the parallel or ensemble topology. The performance of such parallel hybrid classifiers significantly depends on their applied weighting approaches and the accuracy strongly relies on the classifiers' weights. In the literature, several different weighting mechanisms have been developed to yield a higher classification rate, which can be generally categorized into three main categories individual classifiers, averaging-based, and optimization-based algorithms. Although these weighting mechanisms have been commonly and frequently used for parallel hybridization, none of them can guarantee that their obtained classification rate will be optimum. In addition, due to the use of iterative procedures, especially by meta-heuristic optimization-based algorithms, their computational time and cost are always unsatisfactory. In this paper, a linear direct optimal weighting (LDOW) approach is proposed in which it can be generally guaranteed that the optimum classification rate will be directly reached. In this way, the proposed approach can achieve the highest classification rate by the desired computational time and cost and both mentioned limitations of currently-used weighting algorithms can be simultaneously lifted. Empirical results of twenty-three different benchmark data sets from six dissimilar domains indicate that the proposed LDOW approach can yield more classification rate in all data sets and can averagely improve 5.36% the performance of its traditional linear direct non-optimal version with the same components. Furthermore, the proposed LDOW approach can averagely improve 6.87% the classification rate of the averaging weighting algorithms that are similar to the proposed algorithm are direct ones. In addition, the proposed model can even improve 2.80% the classification rate of nonlinear intelligent weighting algorithms, on average. Whereas, the proposed approach is a linear model and its complexity and computational cost are significantly lower than these nonlinear intelligent algorithms. Numerical results also indicate that the proposed model can averagely improve 6.08% the classification rate of meta-heuristic-based weighting algorithms. While the proposed model is a direct model and its computational cost is meaningfully lower than these iterative algorithms. Thus, in theory as well as in practice, it can be inferred that the proposed LDOW approach can be an efficient alternative weighting method for parallel hybridization in the classification field. This is particularly relevant when more accurate results are required or for big data situations where computational time and cost are critical factors to consider. • Proposing a new linear direct optimal weighting (LDOW) algorithm. • Developing a parallel hybrid classification model based on the proposed algorithm. • Illustrating the efficiency of the proposed model in the different domains and data sets. [ABSTRACT FROM AUTHOR]

In today's digital environment, business intelligence advances make it difficult to stay competitive and up to date on business trends. Decision-making in the financial industry is increasingly being powered by big data and machine learning. A decision-making process may be thought of as any sequence of processes that an individual goes through in order to select the option or course of action that is most suitable to meet their needs and necessities. The ability to anticipate the onset of a financial crisis is a significant economic phenomenon. A nation's economic development and strength can be gauged by its capacity to provide an accurate assessment of the number of failed firms and the frequency with which they fail. The economics of the globe have been ravaged by recent global crises like as the COVID-19 pandemic and other recent environmental, financial, and economic disasters, which have marginalized efforts to construct a maintainable economy and civilization. The health and growth of a nation's economy can be determined by precisely estimating the number of enterprises that will fail and the number that will succeed. Historically, there have been numerous strategies for constructing a successful financial crisis prediction (FPC) method. Effectively predicting business failures is a gauge of a country's economic health. Several strategies are available for effective FCP. Classification performance, forecast accuracy, and legality are insufficient for practical use. Several of the suggested methods work for some issues. The specific dataset is not expandable. To improve classification, design a good prediction model adaptable to several datasets. An effective financial crisis prediction method (FPC) requires the right qualities. ML models can also be used to classify a company's financial health. This research presents political optimizer-based feature selection (POFS) with optimal cascaded deep forest (OCDF) for FCP in big data environments. Hadoop Map Reduce handles huge datasets. POFS reduces computing complexity by handling feature selection. POFS is an original FCP algorithm categorization using OCDF. SFO is used to optimize CDF model parameters. A thorough simulation study was performed to evaluate POFS performance on benchmark datasets OCDFs. The results confirmed the POFS-OCDF method's superiority over state-of-the-art approaches. With an outstanding sensitivity of 0.912, specificity of 0.953, accuracy of 0.944, F-score of 0.930, and Matthews correlation coefficient (MCC) of 0.912, the proposed POFS-OCDF technique has shown optimum results. The experimental results demonstrated that the POFS-OCDF technique outperformed other recently developed strategies on a variety of criteria. As previously stated, Sunflower optimization (SFO) is also used to tune the Cascaded Deep Forest (CDF) parameters. A detailed simulation analysis is performed based on the benchmark dataset to evaluate the higher classification efficiency of the POFS-OCDF technique. The invention of the POFS algorithm for FCP exemplifies the work's originality. [ABSTRACT FROM AUTHOR]

Recently, with the development of technology, it is quite important to study scalable computational methods for handling large-scale data in big data applications. The cloud/edge computing are powerful tools for solving big data problems, and provide flexible computation and huge storage capability. Moreover, in the real world, the data in big data applications usually are stored in decentralized computation resources, which affects the design of artificial intelligence algorithms. Therefore, in this work, we focus on distribution machine learning, and propose a novel distributed classification algorithm to deal with imbalanced data. Specifically, we explore the distributed alternating direction method of multipliers (ADMM) framework, and divide the distributed classification problem as some small problems which can be solved by the decentralized resources in parallel. Furthermore, based on the distributed framework, we use a acceleration strategy to improve the time efficiency with designing a more suitable model for imbalanced data classification. The theoretical analysis and experiments results show that our proposed method converges faster than other distributed ADMM method and saves training time, which can improve the scalability of distributed classification on imbalanced data. [ABSTRACT FROM AUTHOR]

INTRODUCTION: The telecommunications industry faces significant challenges due to customer attrition, which directly impacts revenue. To better understand and address this issue, Companies are looking into techniques to determine the internal issues that affect customer churn. OBJECTIVES: This article offers an overview of customer attrition within the telecommunications sector. METHODS: It introduces an advanced churn prediction model harnessing state-of-the-art technologies, including neural networks, machine learning, and other cutting-edge innovations, to achieve remarkably high accuracy rates. By analyzing diverse parameters and datasets collected from multiple telecom companies, valuable insights can be gained. RESULTS: The model's performance on test data can be evaluated using metrics such as Accuracy Score, Area under Curve (AUC), Sensitivity, Specificity, and other performance indicators. CONCLUSION: In order to effectively manage extensive datasets, organizations can leverage Big Data technology. This empowers them to forecast the probability of customer churn and put in place proactive strategies to retain their customer base. [ABSTRACT FROM AUTHOR]

Machine learning techniques that rely on textual features or sentiment lexicons can lead to erroneous sentiment analysis. These techniques are especially vulnerable to domain-related difficulties, especially when dealing in Big data. In addition, labeling is time-consuming and supervised machine learning algorithms often lack labeled data. Transfer learning can help save time and obtain high performance with fewer datasets in this field. To cope this, we used a transfer learning-based Multi-Domain Sentiment Classification (MDSC) technique. We are able to identify the sentiment polarity of text in a target domain that is unlabeled by looking at reviews in a labelled source domain. This research aims to evaluate the impact of domain adaptation and measure the extent to which transfer learning enhances sentiment analysis outcomes. We employed transfer learning models BERT, RoBERTa, ELECTRA, and ULMFiT to improve the performance in sentiment analysis. We analyzed sentiment through various transformer models and compared the performance of LSTM and CNN. The experiments are carried on five publicly available sentiment analysis datasets, namely Hotel Reviews (HR), Movie Reviews (MR), Sentiment140 Tweets (ST), Citation Sentiment Corpus (CSC), and Bioinformatics Citation Corpus (BCC), to adapt multi-target domains. The performance of numerous models employing transfer learning from diverse datasets demonstrating how various factors influence the outputs. [ABSTRACT FROM AUTHOR]

One of the most important tasks in classification is to design a high-precision classifier that has a small training and testing time. Classifier such as support vector machine is an efficient and well-known classifier. However, in the face of large data sets, it has a problem with performance. Another major problem with most data sets is the high boundary complexity among classes, which will reduce generalizability. To address these problems, a novel strategy of large-scale data classification is proposed. A simple and useful method is presented to reconstruct the training data set. The goal is to speed up the classifier with minimal impact on performance. It has three phases: in the first step, we try to select the appropriate representatives for the data; in the second step, the data causing the complexity of the boundary will be removed; in the third step, the classification is performed on the new data set. The proposed method is tested with 23 datasets and comparatively evaluated against five of the most successful instance-based condensation algorithms. Experiments showed that despite the simplicity of the proposed method, its performance is better than other methods presented in the research literature. [ABSTRACT FROM AUTHOR]

The rapid pace of urbanization and increasing demands for urban functionalities have led to diversification and complexity in the types of urban surface elements. The conventional approach of relying solely on remote sensing imagery for urban surface element extraction faces emerging challenges. Data-driven techniques, including deep learning and machine learning, necessitate a substantial number of annotated samples as prerequisites. In response, our study proposes a knowledge-driven approach that integrates multisource data with ontology to achieve precise urban surface element extraction. Within this framework, components from the EIONET Action Group on Land Monitoring in Europe matrix serve as ontology primitives, forming a shared vocabulary. The semantics of surface elements are deconstructed using these primitives, enabling the creation of specific descriptions for various types of urban surface elements by combining these primitives. Our approach integrates multitemporal high-resolution remote sensing data, network big data, and other heterogeneous data sources. It segments high-resolution images into individual patches, and for each unit, urban surface element classification is accomplished through semantic rule-based inference. We conducted experiments in two regions with varying levels of urban scene complexity, achieving overall accuracies of 93.03% and 97.35%, respectively. Through this knowledge-driven approach, our proposed method significantly enhances the classification performance of urban surface elements in complex scenes, even in the absence of sample data, thereby presenting a novel approach to urban surface element extraction. [ABSTRACT FROM AUTHOR]

The proposed system aims to overcome challenges posed by large databases, data imbalance, heterogeneity, and multidimensionality through progressive sampling as a novel classification model. It leverages sampling techniques to enhance processing performance and overcome memory restrictions. The random forest regressor feature importance technique with the gini significance method is employed to identify important characteristics, reducing the data's features for classification. The system utilizes diverse classifiers such as random forest, ensemble learning, support vector machine (SVM), k-nearest neighbors' algorithm (KNN), and logistic regression, allowing flexibility in handling different data types and achieving high accuracy in classification tasks. By iteratively applying progressive sampling to the dataset with the best features, the proposed technique aims to significantly improve performance compared to using the entire dataset. This approach focuses computational resources on the most informative subsets of data, reducing time complexity. Results show that the system can achieve over 85% accuracy even with only 5-10% of the original data size, providing accurate predictions while reducing data processing requirements. In conclusion, the proposed system combines progressive sampling, feature selection using random forest regressor feature importance (RFRFI-PS), and a range of classifiers to address challenges in large databases and improve classification accuracy. It demonstrates promising results in accuracy and time complexity reduction. [ABSTRACT FROM AUTHOR]