Search

Artificial Intelligence is providing astonishing results, with medicine being one of its favourite playgrounds. Machine Learning and, in particular, Deep Neural Networks are behind this revolution. Among the most challenging targets of interest in medicine are cancer diagnosis and therapies but, to start this revolution, software tools need to be adapted to cover the new requirements. In this sense, learning tools are becoming a commodity but, to be able to assist doctors on a daily basis, it is essential to fully understand how models can be interpreted. In this survey, we analyse current machine learning models and other in-silico tools as applied to medicine-specifically, to cancer research-and we discuss their interpretability, performance and the input data they are fed with. Artificial neural networks (ANN), logistic regression (LR) and support vector machines (SVM) have been observed to be the preferred models. In addition, convolutional neural networks (CNNs), supported by the rapid development of graphic processing units (GPUs) and high-performance computing (HPC) infrastructures, are gaining importance when image processing is feasible. However, the interpretability of machine learning predictions so that doctors can understand them, trust them and gain useful insights for the clinical practice is still rarely considered, which is a factor that needs to be improved to enhance doctors' predictive capacity and achieve individualised therapies in the near future.

Introduction: Ovarian tumors are the most common diagnostic challenge for gynecologists and ultrasound examination has become the main technique for assessment of ovarian pathology and for preoperative distinction between malignant and benign ovarian tumors. However, ultrasonography is highly examiner-dependent and there may be an important variability between two different specialists when examining the same case. The objective of this work is the evaluation of different well-known Machine Learning (ML) systems to perform the automatic categorization of ovarian tumors from ultrasound images., Methods: We have used a real patient database whose input features have been extracted from 348 images, from the IOTA tumor images database, holding together with the class labels of the images. For each patient case and ultrasound image, its input features have been previously extracted using Fourier descriptors computed on the Region Of Interest (ROI). Then, four ML techniques are considered for performing the classification stage: K-Nearest Neighbors (KNN), Linear Discriminant (LD), Support Vector Machine (SVM) and Extreme Learning Machine (ELM)., Results: According to our obtained results, the KNN classifier provides inaccurate predictions (less than 60% of accuracy) independently of the size of the local approximation, whereas the classifiers based on LD, SVM and ELM are robust in this biomedical classification (more than 85% of accuracy)., Conclusions: ML methods can be efficiently used for developing the classification stage in computer-aided diagnosis systems of ovarian tumor from ultrasound images. These approaches are able to provide automatic classification with a high rate of accuracy. Future work should aim at enhancing the classifier design using ensemble techniques. Another ongoing work is to exploit different kind of features extracted from ultrasound images., Competing Interests: The authors have declared that no competing interests exist.

We are witnessing the digitalization era, where artificial intelligence (AI)/machine learning (ML) models are mandatory to transform this data deluge into actionable information. However, these models require large, high-quality datasets to predict high reliability/accuracy. Even with the maturity of Internet of Things (IoT) systems, there are still numerous scenarios where there is not enough quantity and quality of data to successfully develop AI/ML-based applications that can meet market expectations. One such scenario is precision agriculture, where operational data generation is costly and unreliable due to the extreme and remote conditions of numerous crops. In this paper, we investigated the generation of synthetic data as a method to improve predictions of AI/ML models in precision agriculture. We used generative adversarial networks (GANs) to generate synthetic temperature data for a greenhouse located in Murcia (Spain). The results reveal that the use of synthetic data significantly improves the accuracy of the AI/ML models targeted compared to using only ground truth data. [ABSTRACT FROM AUTHOR]

Nowadays, human overpopulation is stressing our ecosystems in different ways, agriculture being a critical example as different predictions point towards food shortages in the near future. Accordingly, smart farming is becoming key to the optimization of natural resources so that different crops can be grown efficiently, consuming as few resources as possible. In particular, greenhouses have proved to be an effective way of producing a high volume of vegetables/fruits in a reduced space and within a short time span. Hence, optimizing greenhouse functioning results in less water use and nutrient consumption, less energy use, faster growth, and better product quality. In this article, we carry out an in-depth analysis of different machine learning (ML) models to improve climate control in smart greenhouses. As part of the analysis of the techniques we also considered 3 ways of pre-processing the data, as well as 12-hour and 24-hour forecasting. We focus on forecasting the indoor air temperature of an operational smart greenhouse, i.e. assessing the data anomalies that are inherently present in these environments due to the instability of IoT infrastructures. Several ML models are adapted to time series forecasting to provide an overview of these techniques and to find out which one performs better in this particular scenario. Our results show that, after statistically validating the results, the Random Forest Regression technique gives the best overall result with a mean absolute error of less than 1 degree Celsius. [ABSTRACT FROM AUTHOR]

Portable mid-infrared (MIR) technology is well suited for the provision of detailed and inexpensive information on key soil properties for optimum soil management. This technology requires prior complex multivariate modelling. In this manuscript, we propose an intelligent system approach based on portable MIR spectroscopy data modelled by machine learning techniques to predict total carbon (TC), nitrogen (TN), cation exchange capacity (CEC), clay, silt and exchangeable sodium (Na+) in 458 representative soil samples from Australia. To find the best performing algorithm to be incorporated into the intelligent system, we evaluated the performance of a number of machine learning techniques, including Gaussian process regression (GPR), Random Forest (RF), M5 rules, and Bagging and Decision Trees, which were compared with traditional Partial Least Squares Regression (PLSR). Generally, all the models showed a satisfactory performance. However, GPR stood out (according to statistical tests), especially for the predictions of clay, Na+ and CEC, obtaining root mean squared error differences between the GPR and PLSR techniques of −0.27%, −0.40% and −0.63 cmol+ kg−1, respectively. This study evidenced the feasibility of using the GPR technique to model data provided by a portable MIR instrument for the prediction of soil properties, providing the best, most stable and reliable results. Highlights • Different machine learning techniques have been tested to predict soil properties. • The predicted soil properties are TC, TN, CEC, clay, silt and Na+. • The best predictive machine learning technique has been the Gaussian Process. • The Gaussian process is better compared to the traditional PLSR technique. • The Gaussian Process is the candidate for the development of intelligent system. [ABSTRACT FROM AUTHOR]

The numerical value discretization is a process that is performed in the data preprocessing phase of intelligent data analysis. Preprocessing phase is very relevant because the quality of the models obtained in data mining step depends on this phase. Value discretization is an important task in data preprocessing because not all data mining techniques can handle continuous values. In this paper an unsupervised technique to discretize continuous data values using fuzzy partitions is proposed. Specifically a clustering technique that gets fuzzy partitions is presented. In addition, to evaluate the behavior of the proposed technique a series of experiments have been proposed using a Extreme Learning Machine classifier and a committee of Extreme Learning Machine. Beside comparing with the K-means discretization technique. These experiments have been validated statistically obtaining the best results the approach proposed. [ABSTRACT FROM AUTHOR]

Faced with a new concept to learn, our brain does not work in isolation. It uses all previously learned knowledge. In addition, the brain is able to isolate the knowledge that does not benefit us, and to use what is actually useful. In machine learning, we do not usually benefit from the knowledge of other learned tasks. However, there is a methodology called Multitask Learning (MTL), which is based on the idea that learning a task along with other related tasks produces a transfer of information between them, what can be advantageous for learning the first one. This paper presents a new method to completely design MTL architectures, by including the selection of the most helpful subtasks for the learning of the main task, and the optimal network connections. In this sense, the proposed method realizes a complete design of the MTL schemes. The method is simple and uses the advantages of the Extreme Learning Machine to automatically design a MTL machine, eliminating those factors that hinder, or do not benefit, the learning process of themain task. This architecture is unique and it is obtained without testing/error methodologies that increase the computational complexity. The results obtained over several real problems show the good performances of the designed networks with this method. [ABSTRACT FROM AUTHOR]

Abstract: Selection of the optimal neural architecture to solve a pattern classification problem entails to choose the relevant input units, the number of hidden neurons and its corresponding interconnection weights. This problem has been widely studied in many research works but their solutions usually involve excessive computational cost in most of the problems and they do not provide a unique solution. This paper proposes a new technique to efficiently design the MultiLayer Perceptron (MLP) architecture for classification using the Extreme Learning Machine (ELM) algorithm. The proposed method provides a high generalization capability and a unique solution for the architecture design. Moreover, the selected final network only retains those input connections that are relevant for the classification task. Experimental results show these advantages. [Copyright &y& Elsevier]

Abstract: In Multi-Task Learning (MTL), when several related tasks are learned at the same time considering one of them as the main task and the others as secondary ones, there is a transfer of positive information that improves the performance of the main one. However, not only does the difficulty of finding the relationship among different tasks pose a problem in real applications, but also knowing the kind of relationship among them. This paper presents a new method to generate artificial hints (subsets from the original data set) that helps the learning of the main task when all of them are learned simultaneously (as in a MTL scheme). Thus, although these hints cannot be strictly considered as secondary tasks, they will act as guides for the main one. The results obtained with toy and real problems show the advantages of the proposed method. In particular, a faster convergence, a very good performance, and a reduction in the likelihood of being trapped in a local minimum are achieved. [Copyright &y& Elsevier]

The availability of water resources is limited, and rising consumption has increased pressure on natural resources. Therefore, reclaimed water represents an alternative option for use in urban areas, industry and, in particular, agriculture. Recent research has shown that some pharmaceutical compounds are not fully removed by wastewater treatment plants (WWTPs) and may eventually be released into agricultural systems through the application of wastewater-based resources (sludge and effluent). The present study develops an intelligent expert system (based on a feedforward neural network trained via an extreme learning machine algorithm) for predicting the carbamazepine (CBZ) and diclofenac (DCF) content in lettuce tissues irrigated with reclaimed water from WWTPs. This reduces laboratory costs, mitigates the negative impacts on the environment and leads to more effective, safer decisions on the use of reclaimed water in agriculture. The results obtained, which were validated through statistical testing, demonstrate that the intelligent expert system is well calibrated and reliable. Finally, this system was used to predict maps of CBZ and DCF accumulation in lettuce crops if they were watered with effluent from 10 WWTPs located in the Region of Murcia (Spain). In conclusion, our system provides highly accurate predictions of the amount of CBZ and DCF contained in different lettuce tissues (roots and leaves) and the predicted concentrations do not present any health risk. Highlights • Expert system predicts pharmaceutical absorption in lettuce by analysing effluent. • Predictions for accumulation in lettuces irrigated from 10 treatment plants in Murcia. • Predicted concentrations in lettuce do not constitute a potential risk to health. [ABSTRACT FROM AUTHOR]

A correct estimation of the instantaneous peak flow (IPF) is crucial to reducing the consequences of flash floods. An approach to estimate the IPF, obtained by combining Soil and Water Assessment Tool (SWAT) simulation and machine-learning models, was proposed and then verified by comparison with observation-based results in the Ladra river basin, northwest Spain. The SWAT model has been used to estimate the maximum mean daily flow (MMDF), and machine-learning models have been used to estimate the IPF based on MMDF. Four nonlinear time-series intelligence models, artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS), support vector machine (SVM) and extreme learning machine (ELM) were applied, and their results were compared. The Modified Nash-Sutcliffe efficiency coefficient (MNSE) and the index of agreement (d) were used to evaluate SWAT performance while simulating MMDF, and the coefficient of determination (R2) and the root mean square error (RMSE) were employed to evaluate the performance of these intelligent systems. According to the results, the SWAT hydrological model is a useful tool to simulate MMDF. Validation analyses resulted in values of statistical indexes (MNSE = 0.64 and d = 0.95). Regarding intelligent systems, the results show that they can be successfully used in predicting IPF, but ELM has demonstrated a superior ability to estimate IPF from the MMDF (R2 = 0.86 and RMSE = 48.59). The results of this study can contribute to predicting IPF in areas where sub-daily observational data are scarce, thereby reducing uncertainties associated with IPF estimations. Graphical abstract Highlights • SWAT model has been used to estimate maximum mean daily flows. • Intelligent Systems estimate instantaneous peak flow based on the daily flows. • Extreme Learning and Support Vector Machine were optimal to estimate peak flow. • Predictions correlated with experimental data with coefficients of 0.86–0.88. • Results assist predicting peak flow with limited sub-daily observational data. [ABSTRACT FROM AUTHOR]

Various methods are used to make the partition of data sets for QSAR development and model validation. In this work we used a fuzzy minimals partitioning and we compare this methodology with another rational partition methods like k -means clustering (KMS) and Minimal Test Set Dissimilarity (MTSD). For the development of QSAR models Ordinary Least Squares (OLS) and Extreme Learning Machine (ELM) methods were used. The generated QSAR equations were validated by the coefficient of determination of the internal leave one out (LOO) cross validation method Q LOO 2 and then the coefficient of the external test set Q ext 2 was compared between partition methods. The results of this comparison showed that using fuzzy minimal for big and structurally diverse data sets gave an applicability domain similar to KMS and a better predictability models than both methods, KMS and MTSD. [ABSTRACT FROM AUTHOR]