Visual defect detection is a critical component of industrial quality control, particularly for metal surfaces, where accurate classification and localization of defects can significantly affect product reliability and aesthetics. However, most deep learning models require large labeled datasets, which are often unavailable in real-world settings. This paper presents a hybrid approach to overcome data scarcity by integrating synthetically generated data into the training pipeline of a state-of-the-art classifier, SuperSimpleNet. We utilize DefectGAN, a generative adversarial network, to create realistic defect images and corresponding pixel-level masks, enhancing the diversity and generalization capacity of the training data. Additionally, we employ custom geometric data augmentation techniques to further enrich the dataset. Experimental results demonstrate significant performance improvements in both image-and pixel-level defect detection metrics. Our findings support the use of generative and augmented data as effective strategies to train high-performance defect detection models under limited data conditions.

Music evokes emotion in many people. We introduce a novel way to manipulate the emotional content of a song using AI tools. Our goal is to achieve the desired emotion while leaving the original melody as intact as possible. For this, we create an interactive pipeline capable of shifting an input song into a diametrically opposed emotion and visualize this result through Russel's Circumplex model. Our approach is a proof-of-concept for Semantic Manipulation of Music, a novel field aimed at modifying the emotional content of existing music. We design a deep learning model able to assess the accuracy of our modifications to key, SoundFont instrumentation, and other musical features. The accuracy of our model is inline with the current state of the art techniques on the 4Q Emotion dataset. With further refinement, this research may contribute to on-demand custom music generation, the automated remixing of existing work, and music playlists tuned for emotional progression.

This paper explores the current landscape of visualizing large language models (LLMs). The main objective was threefold. Firstly, we investigate how we can visualize LLM-specific techniques such as prompt engineering, instruction tuning, or guidance. Secondly, LLM causality, interpretability, and explainability are examined through visualization. And finally, we showcase the role of visualization in illuminating the integration of multiple modalities. We are interested in discovering the papers that present visualization systems instead of those that use visualization to showcase a part of their work. Our survey aims to synthesize the state-of-the-art in LLM visualization, offering a compact resource for exploring future research avenues.

In machine learning, concept drift is an evolution of information that invalidates the current data model. It happens when the statistical properties of the input data change over time in unforeseen ways. Concept drift detection is crucial when dealing with dynamically changing data. Its visualization can bring valuable insight into the data dynamics, especially for multidimensional data, and is related to visual knowledge discovery. We propose a novel visualization model based on parallel coordinates, denoted as parallel histograms through time. Our model represents histograms of feature distributions for successive time-shifted windows. The drift is shown as variations of these histograms, obtained by connecting the means of the distribution for successive time windows. We show how these diagrams can be used to explain the decision made by the machine learning model in choosing the drift point. By isolating the drift at the edges of successive time windows, there will be none (or reduced) drift within the adjacent windows. We illustrate this concept on both synthetic and real datasets. In our experiments, we use an incremental/decremental SVM with shifting window, introduced by us in previous work. With our proposed technique, in addition to detect the presence of concept drift, we can also depict it. This information can be further used to explain the change.

Graph Convolutional Networks (GCN) are Graph Neural Networks where the convolutions are applied over a graph. In contrast to Convolutional Neural Networks, GCN's are designed to perform inference on graphs, where the number of nodes can vary, and the nodes are unordered. In this study, we address two important challenges related to GCNs: i) oversmoothing; and ii) the utilization of node relational properties (i.e., heterophily and homophily). Oversmoothing is the degradation of the discriminative capacity of nodes as a result of repeated aggregations. Heterophily is the tendency for nodes of different classes to connect, whereas homophily is the tendency of similar nodes to connect. We propose a new strategy for addressing these challenges in GCNs based on Transfer Entropy (TE), which measures of the amount of directed transfer of information between two time varying nodes. Our findings indicate that using node heterophily and degree information as a node selection mechanism, along with feature-based TE calculations, enhances accuracy across various GCN models. Our model can be easily modified to improve classification accuracy of a GCN model. As a trade off, this performance boost comes with a significant computational overhead when the TE is computed for many graph nodes.

In a feedforward network, Transfer Entropy (TE) can be used to measure the influence that one layer has on another by quantifying the information transfer between them during training. According to the Information Bottleneck principle, a neural model's internal representation should compress the input data as much as possible while still retaining sufficient information about the output. Information Plane analysis is a visualization technique used to understand the trade-off between compression and information preservation in the context of the Information Bottleneck method by plotting the amount of information in the input data against the compressed representation. The claim that there is a causal link between information-theoretic compression and generalization, measured by mutual information, is plausible, but results from different studies are conflicting. In contrast to mutual information, TE can capture temporal relationships between variables. To explore such links, in our novel approach we use TE to quantify information transfer between neural layers and perform Information Plane analysis. We obtained encouraging experimental results, opening the possibility for further investigations.

In recent years, pruning has emerged as a popular technique to reduce the computational complexity and memory footprint of Convolutional Neural Network (CNN) models. Mutual Information (MI) has been widely used as a criterion for identifying unimportant filters to prune. However, existing methods for MI computation suffer from high computational cost and sensitivity to noise, leading to suboptimal pruning performance. We propose a novel method to improve MI computation for CNN pruning, using the spatial aura entropy. The spatial aura entropy is useful for evaluating the heterogeneity in the distribution of the neural activations over a neighborhood, providing information about local features. Our method effectively improves the MI computation for CNN pruning, leading to more robust and efficient pruning. Experimental results on the CIFAR-10 benchmark dataset demonstrate the superiority of our approach in terms of pruning performance and computational efficiency.

The quantum circuit layout (QCL) problem involves mapping out a quantum circuit such that the constraints of the device are satisfied. We introduce a quantum circuit mapping heuristic, QXX, and its machine learning version, QXX-MLP. The latter automatically infers the optimal QXX parameter values such that the laid out circuit has a reduced depth. In order to speed up circuit compilation, before laying the circuits out, we use a Gaussian function to estimate the depth of the compiled circuits. This Gaussian also informs the compiler about the circuit region that influences most the resulting circuit’s depth. We present empiric evidence for the feasibility of learning the layout method using approximation. QXX and QXX-MLP open the path to feasible large-scale QCL methods.

The introduction of Transformer neural networks has changed the landscape of Natural Language Processing (NLP) during the recent years. These models are very complex, and therefore hard to debug and explain. In this context, visual explanation became an attractive approach. The visualization of the path that leads to certain outputs of a model is at the core of visual explanation, as this illuminates the features or parts of the model that may need to be changed to achieve the desired results. In particular, one goal of a NLP visual explanation is to highlight the most significant parts of the text that have the greatest impact on the model output. Several visual explanation methods for NLP models were recently proposed. A major challenge is how to compare the performances of such methods since we cannot simply use the usual classification accuracy measures to evaluate the quality of visualizations. We need good metrics and rigorous criteria to measure how useful the extracted knowledge is for explaining the models. In addition, we want to visualize the differences between the knowledge extracted by different models, in order to be able to rank them. In this paper, we investigate how to evaluate explanations/visualizations resulted from machine learning models for text classification. The goal is not to improve the accuracy of a particular NLP classifier, but to assess the quality of the visualizations that explain its decisions. We describe several methods for evaluating the quality of NLP visualizations, including both automated techniques based on quantifiable measures and subjective techniques based on human judgements.

Visual interpretation of Synthetic Aperture Radar (SAR) images plays an important role in remote sensing mainly because SAR images enable consistent monitoring in any lighting, weather, and cloud-cover conditions. An important application of SAR visualization is water detection. We introduce a Fully Convolutional Neural Network (FCN) Encoder to detect water in Sentinel-1 SAR images. Our FCN Encoder identifies water by the intensity of each pixel and also learns the spatial information of neighborhood pixels. We apply our method on standard benchmarks and real-world SAR images. The results are assessed both visually and from the point of view of classification accuracy. Compared with other classifiers, our FCN Encoder is more accurate. From visual inspection of the Seattle water detection result, the FCN Encoder produces a very clear (smooth) output. The results show that the FCN Encoder, trained with a harder dataset and hyperparameter optimization, improves significantly its generalization performance. In a real-world application, for the prediction phase, the FCN Encoder is about 40 times faster than a Convolutional Neural Network (CNN) with sliding window.

We study the problem of learning the data samples' distribution as it changes in time. This change, known as concept drift, complicates the task of training a model, as the predictions become less and less accurate. It is known that Support Vector Machines (SVMs) can learn weighted input instances and that they can also be trained online (incremental-decremental learning). Combining these two SVM properties, the open problem is to define an online SVM concept drift model with shifting weighted window. The classic SVM model should be retrained from scratch after each window shift. We introduce the Weighted Incremental-Decremental SVM (WIDSVM), a generalization of the incremental-decremental SVM for shifting windows. WIDSVM is capable of learning from data streams with concept drift, using the weighted shifting window technique. The soft margin constrained optimization problem imposed on the shifting window is reduced to an incremental-decremental SVM. At each window shift, we determine the exact conditions for vector migration during the incremental-decremental process. We perform experiments on artificial and real-world concept drift datasets; they show that the classification accuracy of WIDSVM significantly improves compared to a SVM with no shifting window. The WIDSVM training phase is fast, since it does not retrain from scratch after each window shift.

The introduction of Transformer neural networks has changed the landscape of Natural Language Processing (NLP) during the last years. So far, none of the visualization systems has yet managed to examine all the facets of the Transformers. This gave us the motivation of the current work. We propose a novel NLP Transformer context-sensitive visualization method that leverages existing NLP tools to find the most significant groups of tokens (words) that have the greatest effect on the output, thus preserving some context from the original text. The original contribution is a context-aware visualization method of the most influential word combinations with respect to a classifier. This context-sensitive approach leads to heatmaps that include more of the relevant information pertaining to the classification, as well as more accurately highlighting the most important words from the input text. The proposed method uses a dependency parser, a BERT model, and the leave-n-out technique. Experimental results suggest that improved visualizations increase the understanding of the model, and help design models that perform closer to the human level of understanding for these problems.

The introduction of Transformer neural networks has changed the landscape of Natural Language Processing during the last three years. While models inspired by it have managed to lead the boards for a variety of tasks, some of the mechanisms through which these performances were achieved are not necessarily well-understood. Our survey is focused mostly on explaining Transformer architectures through visualizations. Since visualization enables some degree of explainability, we have examined the various Transformer facets that can be explored through visual analytics. The field is still at a nascent stage and is expected to witness dynamic growth in the near future, since the results are already interesting and promising. Currently, some of the visualizations are relatively close to their original models, whereas others are model-agnostic. The visualizations designed to explore the Transformer architectures enable some additional features, like exploration of all neuronal cells or attention maps, therefore providing an advantage for this particular task. We conclude by proposing a set of requirements for future Transformer visualization frameworks.

The classification of stellar spectra is a fundamental task in stellar astrophysics. There have been many explorations into the automated classification of stellar spectra but few that involve the Sloan Digital Sky Survey (SDSS). We use the SDSS dataset since it is the most important stellar spectra database available today. In our approach, we apply redshift corrections to the spectra and reduce the number of flux measurements by feature selection. Then we apply standard classifier methods: Random Forest and Support Vector Machine. We compare the accuracy of feature selection and classifier combinations for redshifted stellar spectra and rest stellar spectra. Even though redshifted stellar spectra create feature matrix discrepancies, classifiers utilizing redshifted stellar spectra perform with high accuracy. This creates a viable option for automated classification of stellar spectra without having to identify the redshift value.

We present the first algorithm for finding holes in high dimensional data that runs in polynomial time with respect to the number of dimensions. Previous algorithms are exponential. Finding large empty rectangles or boxes in a set of points in 2D and 3D space has been well studied. Efficient algorithms exist to identify the empty regions in these low-dimensional spaces. Unfortunately such efficiency is lacking in higher dimensions where the problem has been shown to be NP-complete when the dimensions are included in the input. Applications for algorithms that find large empty spaces include big data analysis, recommender systems, automated knowledge discovery, and query optimization. Our Monte Carlo-based algorithm discovers interesting maximal empty hyper-rectangles in cases where dimensionality and input size would otherwise make analysis impractical. The run-time is polynomial in the size of the input and the number of dimensions. We apply the algorithm on a 39-dimensional data set for protein structures and discover interesting properties that we think could not be inferred otherwise.

Our research area is the unilateral dependency (UD) analysis of non-linear relationships within pairs of simultaneous data. The application is in financial analysis, using the data reported by Kodak and Apple for the period of 1999-2014. We compute and analyze the UD between Kodak's and Apple's financial time series in order to understand how they influence each other over their company assets and liabilities. We also analyze within each of the two companies the UD between assets and liabilities. Our formal approach is based on the informational energy UD measure derived by us in previous work. This measure is estimated here from available sample data, using a non-parametric asymptotically unbiased and consistent kNN estimator.

We propose a method for music genre classification based on a Self-Organizing Map (SOM) - type network. Music pieces are viewed as sequences of pitch and timbre signals. We define a similarity measure between these sequences, derived from the Levenshtein (edit) distance. In contrast to the standard Levenshtein distance, our similarity measure is able to operate on a continuous vector space. Using this measure, we map the input music pieces on a SOM. The SOM is trained using a special string adjustment mechanism, which is determined by an algebraic equation. Our method turns out to achieve better classification accuracy than some other recent techniques. The feature set identified by SOM provides superior classifier accuracy compared to the same classifier applied on a random feature set of the same size. On standard benchmarks, two of our derived classifiers achieve accuracies of 97.32% (using a slow kNN learning algorithm), respectively 95.20% (using a SOM - type algorithm).

Contemporary sensor data are generally large data streams, possibly at a high sampling rate, making data analysis and visualization complex and computationally intensive. We present a novel clustering method for the evaluation of signal data. We are interested in clustering the signals based on the similarity of their behavior (shape), which contains more information than the signal intensity and the dominant frequencies. The signals are encoded into symbol strings. We use the edit distance to determine the similarity between strings. Based on this similarity, we cluster the data streams into a SOM-type network. This SOM is dynamic and adapts incrementally to the input sensor data stream. Incoming signals are processed on the fly and the system has the capability to “forget” old signals. Our method is particularly useful for the inspection of signal streams, both in the context of on-line monitoring and off-line analysis, and can be used as a component in a visualization dashboard.

Fast and accurate determination of the 3D structure of molecules is essential for better understanding their physical, chemical, and biological properties. We focus on an existing method for molecular structure determination: restrained molecular dynamics with simulated annealing. In this method a hybrid function, composed by a physical model and experimental restraints, is minimized by simulated annealing. Our goal is to accelerate computation time using commodity multi-core CPUs and GPUs in a heterogeneous computing model. We present a parallel and portable OpenCL implementation of this method. Experimental results are discussed in terms of accuracy, execution time, and parallel scalability. With respect to the XPLOR-NIH professional software package, compared to the single CPU core implementation, we obtain speedups of three to five times (increasing with problem size) on commodity GPUs. We achieve these performances by writing specialized kernels for different problem sizes and hardware architectures.

Several neural architectures were successfully used to predict properties of chemical compounds. Obtaining satisfactory results with neural networks depends on the availability of large data samples. However, most classical Quantitative Structure-Activity Relationship studies have been performed on small datasets. Neural models do generally infer with difficulty from such datasets. In our study, we analyze the performance of the Bayesian ARTMAP for the prediction of biological activities of HIV-1 protease inhibitors, when inferring from a small and structurally diverse dataset of molecules. The Bayesian ARTMAP is a neural model which uses both competitive learning and Bayesian prediction, and has both the universal approximation and best approximation properties. It is the first time when this model is used in a “real-world” function approximation application. We compare the performance of the Bayesian ARTMAP to several other models, each implementing a different learning mechanism. Experiments are performed within Weka's “Experimenter” standard environment. For our small and structurally diverse dataset of chemical compounds, the Bayesian ARTMAP is a good prediction tool, and the most accurate prediction models are the ones which perform local approximation.

The informational energy (IE) can be interpreted as a measure of average certainty. In previous work, we have introduced a non-parametric asymptotically unbiased and consistent estimator of the IE. Our method was based on the kth nearest neighbor (kNN) method, and it can be applied to both continuous and discrete spaces, meaning that we can use it both in classification and regression algorithms. Based on the IE, we have introduced a unilateral dependency measure between random variables. In the present paper, we show how to estimate this unilateral dependency measure from an available sample set of discrete or continuous variables, using the kNN and the naïve histogram estimators. We experimentally compare the two estimators. Then, in a real-world application, we apply the kNN and the histogram estimators to approximate the unilateral dependency between random variables which describe the temperatures of sensors placed in a refrigerating room.

The Pacific Northwest Geodesic Array at Central Washington University collects telemetered streaming data from 450 GPS stations. These real-time data are used to monitor and mitigate natural hazards arising from earthquakes, volcanic eruptions, landslides, and coastal sea-level hazards in the Pacific Northwest. Recent improvements in both accuracy of positioning measurements and latency of terrestrial data communication have led to the ability to collect data with higher sampling rates. For seismic monitoring applications, this means 1350 separate position streams from stations located across 1200 km along the West Coast of North America must be able to be both visually observed and automatically analyzed at a sampling rate of up to 1 Hz. Our goal is to efficiently extract and visualize useful information from these data streams. We propose a method to visualize the geodetic data by clustering the signal types with a Self-Organizing Map (SOM). The similarity measure in the SOM is determined by the similarity of signals received from GPS stations. Signals are transformed to symbol strings, and the distance measure in the SOM is defined by an edit distance. The symbol strings represent data streams and the SOM is dynamic. We overlap the resulted dynamic SOM on the Google Maps representation.

We present a combination of methods addressing the molecular distance problem, implemented on a graphic processing unit. First, we use geometric build-up and depth-first graph traversal. Next, we refine the solution by simulated annealing. For an exact but sparse distance matrix, the build-up method reconstructs the 3D structures with a root-mean-square error (RMSE) in the order of 0.1 Å. Small and medium structures (up to 10,000 atoms) are computed in less than 10 seconds. For the largest structures (up to 100,000 atoms), the build-up RMSE is 2.2 Å and execution time is about 540 seconds. The performance of our approach depends largely on the graph structure. The SA step improves accuracy of the solution to the expense of a computational overhead.

A challenging problem in peer-to-peer (P2P) networks is the management of super peers. We understand by this how to dynamically adapt the network topology (the number and the locations of super peers) in accordance to the network changes. The super peers are cluster centers which dynamically adapt their number and location. We introduce a self-organizing super peer overlay that suits the communication requirements of a P2P system. Our approach is based on the Growing Neural Gas clustering algorithm. The proposed framework may be suitable for disseminating network services in dynamic and large-scale networks where a large number of data and services need to be replicated, moved, and deleted in a decentralized manner. In our experiments, performed on the Protopeer simulator, the proposed algorithm adapts well to variable network load and churn.

Obtaining satisfactory results with neural networks depends on the availability of large data samples. The use of small training sets generally reduces performance. Most classical Quantitative Structure-Activity Relationship (QSAR) studies for a specific enzyme system have been performed on small data sets. We focus on the neuro-fuzzy prediction of biological activities of HIV-1 protease inhibitory compounds when inferring from small training sets. We propose two computational intelligence prediction techniques which are suitable for small training sets, at the expense of some computational overhead. Both techniques are based on the FAMR model. The FAMR is a Fuzzy ARTMAP (FAM) incremental learning system used for classification and probability estimation. During the learning phase, each sample pair is assigned a relevance factor proportional to the importance of that pair. The two proposed algorithms in this paper are: 1) The GA-FAMR algorithm, which is new, consists of two stages: a) During the first stage, we use a genetic algorithm (GA) to optimize the relevances assigned to the training data. This improves the generalization capability of the FAMR. b) In the second stage, we use the optimized relevances to train the FAMR. 2) The Ordered FAMR is derived from a known algorithm. Instead of optimizing relevances, it optimizes the order of data presentation using the algorithm of Dagher et al. In our experiments, we compare these two algorithms with an algorithm not based on the FAM, the FS-GA-FNN introduced in [4], [5]. We conclude that when inferring from small training sets, both techniques are efficient, in terms of generalization capability and execution time. The computational overhead introduced is compensated by better accuracy. Finally, the proposed techniques are used to predict the biological activities of newly designed potential HIV-1 protease inhibitors.

We focus on extracting rules from a trained FAMR model. The FAMR is a Fuzzy ARTMAP (FAM) incremental learning system used for classification, probability estimation, and function approximation. The set of rules generated is post-processed in order to improve its generalization capability. Our method is suitable for small training sets. We compare our method with another neuro-fuzzy algorithm, and two standard decision tree algorithms: CART trees and Microsoft Decision Trees. Our goal is to improve efficiency of drug discovery, by providing medicinal chemists with a predictive tool for bioactivity of HIV-1 protease inhibitors.

The Fuzzy ARTMAP with Relevance factor (FAMR) is a Fuzzy ARTMAP (FAM) neural architecture with the fol- lowing property: Each training pair has a relevance factor assigned to it, proportional to the importance of that pair during the learning phase. Using a relevance factor adds more flexibility to the training phase, allowing ranking of sample pairs according to the confidence we have in the in- formation source. We focus on the prediction of biological activities of HIV- 1 protease inhibitory compounds, both known and novel, using a FAMR model. Our new approach consists of two stages: i) During the first stage, we use a genetic algo- rithm (GA) to optimize the relevances assigned to the train- ing data. This improves the generalization capability of the FAMR. ii) In the second stage we use the optimized rele- vances to train the FAMR. Finally, the trained FAMR is used to predict the biological activities of newly designed poten- tial HIV-1 protease inhibitors.

We address the problem of optimizing the distribution of partially replicated databases over a computer network. Replication is used to increase data availability in the presence of site or communication failures and to decrease retrieval costs by local access if possible. We present a new bio-inspired replication management approach which is adaptive, completely decentralized, and based on swarm intelligence. Each node has the autonomy to start at any time, depending on the internal state of its stored data objects, a redistribution process. "Redistribution" means replicate, create, delete, update, or move data objects to other nodes of the network. The redistribution process is a dynamic load-balancing scheme which runs with lower priority in the background. The system is event-driven, but the learning process is not synchronized with the events.

Using a neural network-fuzzy logic-genetic algorithm approach we generate an optimal predictor for biological activities of HIV-1 protease potential inhibitory compounds. We use genetic algorithms (GAs) in the two optimization stages. In the first stage, we generate an optimal subset of features. In the second stage, we optimize the architecture of the fuzzy neural network. The optimized network is trained and used for the prediction of biological activities of newly designed chemical compounds. Finally, we extract fuzzy IF/THEN rules. These rules map physico-chemical structure descriptors to predicted inhibitory values. The optimal subset of features, combined with the generated rules, can be used to analyze the influence of descriptors.

The focus of this study is how we can efficiently implement the neural network backpropagation algorithm on a network of computers (NOC) for concurrent execution. We assume a distributed system with heterogeneous computers and that the neural network is replicated on each computer. We propose an architecture model with efficient pattern allocation that takes into account the speed of processors and overlaps the communication with computation. The training pattern set is distributed among the heterogeneous processors with the mapping being fixed during the learning process. We provide a heuristic pattern allocation algorithm minimizing the execution time of backpropagation learning. The computations are overlapped with communications. Under the condition that each processor has to perform a task directly proportional to its speed, this allocation algorithm has polynomial-time complexity. We have implemented our model on a dedicated network of heterogeneous computers using Sejnowski's NetTalk benchmark for testing. Copyright © 2005 John Wiley & Sons, Ltd.

We introduce a new fuzzy ARTMAP (FAM) neural network: Fuzzy ARTMAP with relevance factor (FAMR). The FAMR architecture is able to incrementally "grow" and to sequentially accommodate input-output sample pairs. Each training pair has a relevance factor assigned to it, proportional to the importance of that pair during the learning phase. The relevance factors are user-defined or computed. The FAMR can be trained as a classifier and, at the same time, as a nonparametric estimator of the probability that an input belongs to a given class. The FAMR probability estimation converges almost surely and in the mean square to the posterior probability. Our theoretical results also characterize the convergence rate of the approximation. Using a relevance factor adds more flexibility to the training phase, allowing ranking of sample pairs according to the confidence we have in the information source. We analyze the FAMR capability for mapping noisy functions when training data originates from multiple sources with known levels of noise.

In this paper we use the maximization of Onicescu's informational energy as a criteria for computing the relevances of input features. This adaptive relevance determination is used in combination with the neural gas and the generalized relevance LVQ algorithms. The idea of applying the neural gas neighborhood cooperation technique to improve the generalized relevance LVQ is due to Hammer et al. and is best described in Hammer et al., 2005. Our approach gives an alternative way for determining the relevances in Hammers's algorithm, and in our experiments it shows at least the same performances. Our contribution is an incremental learning algorithm for supervised classification and feature ranking.

Input feature ranking and selection represent a necessary preprocessing stage in classification, especially when one is required to manage large quantities of data. We introduce a weighted generalized LVQ algorithm, called energy generalized relevance LVQ (EGRLVQ), based on the Onicescu's informational energy. EGRLVQ is an incremental learning algorithm for supervised classification and feature ranking.

An incremental, nonparametric probability estimation procedure using a variation of the Fuzzy ARTMAP (FAM) neural network is introduced. The resulted network, called Fuzzy ARTMAP with relevance factor (FAMR), uses a relevance factor assigned to each sample pair, proportional to the importance of the respective pair during the learning phase. Experimental results have shown that FAMR favorably compares with FAM and probabilistic FAM, both as a classifier and as a probability estimator.

Distributed Computing Systems are a viable and less expensive alternative to parallel computers. However, a serious difficulty in concurrent programming of a distributed system is how to deal with scheduling and load balancing of such a system which may consist of heterogeneous computers. Distributed scheduling schemes suitable for parallel loops with independent iterations on heterogeneous computer clusters have been designed in the past. In this work we consider a class of Self-Scheduling schemes for parallel loops with independent iterations which have been applied to multiprocessor systems. We extend this type of schemes to heterogeneous distributed systems. We present tests that the distributed versions of these schemes maintain load balanced execution on heterogeneous systems.