In computational pathology, the gigapixel scale of Whole-Slide Images (WSIs) requires their decomposition into thousands of patches, resulting in high-dimensional embeddings that are computationally costly to process and often dominated by uninformative regions. Existing patch selection methods typically rely on heuristic sampling and do not explicitly address the trade-off between representation compactness and diagnostic accuracy. To address this gap, we propose EvoPS (Evolutionary Patch Selection), a novel framework that formulates patch selection within the training embedding space as a multi-objective optimization problem and leverages an evolutionary search to simultaneously minimize the number of selected patch embeddings and maximize the performance of a downstream similarity search task, generating a Pareto front of optimal trade-off solutions. By identifying a compact and diagnostically informative subset of training patches, EvoPS produces higher-quality training representations that reduce memory requirements and improve the signal-to-noise ratio of the training set. We validated our framework across four major cancer cohorts from The Cancer Genome Atlas (TCGA) using five histopathology foundation models. The results demonstrate that EvoPS can reduce the required number of training patches by over 90% while consistently maintaining or even improving the final classification F 1 -score compared to a state-of-the-art patch selection method. The EvoPS framework provides a robust and principled method for creating efficient, accurate, and interpretable WSI representations, empowering users to select an optimal balance between computational cost and diagnostic performance.

Digital pathology has transformed cancer diagnosis through computational analysis of whole slide images (WSIs). However, the high-dimensional nature of extracted features presents a major challenge for accurate classification, necessitating an effective feature weighting strategy. In this work, we propose a novel two-phase multi-objective feature weighting framework for cancer classification. In the first phase, we introduce a new frequency-based feature selection method to reduce the search space by eliminating redundant and irrelevant features. In the second phase, we apply a novel feature weighting approach that integrates feature correlation in the initialization process, increasing the likelihood of selecting highly correlated features together. Additionally, features with weights below a predefined threshold are zeroed out, further refining the selection while enhancing classification accuracy. Our experimental results on features extracted from TCGA histopathology images show that the proposed approach outperforms direct feature weighting on all features. By leveraging correlation-aware initialization and frequency-based selection, our method effectively optimizes feature representation, leading to improved cancer classification performance in digital pathology.

The emergence of bias in deep neural models represents a significant reliability concern, which may lead to overoptimistic results on seen data while compromising the model's ability to generalize effectively on unseen datasets. Recent studies conducted on The Cancer Genome Atlas (TCGA), which is a publicly available repository of histopathology images, reveal that the TCGA cancerous features extracted by deep neural networks surprisingly are able to discriminate slides based on their origin sites. This finding undoubtedly indicates the existence of site-specific patterns embedded in the extracted features learned by deep networks rather than focusing on histomorphologic patterns. Consequently, this biased behavior raises concerns about the reliability of these networks. This observation motivates us to conduct a series of experiments in which we present two distinct evolutionary feature selection strategies, each differentiated by its objective function evaluation. The primary goal is to select the features with a minimized foot-print of data source signatures, thereby ensuring a more accurate and site-independent cancer classification. We have conducted nine comprehensive independent experiments across nine cancer types, employing each evolutionary strategy. The comparison between results obtained through evolutionary strategies and the original feature sets highlights the substantial impact of feature selection methods on bias reduction while maintaining accuracy in cancer-type discrimination. Furthermore, the comparison of the two strategies with each other demonstrates the intricate nature of bias and its integration with cancerous features during the training process.

Feature selection plays a pivotal role in the data preprocessing and model-building pipeline, significantly en-hancing model performance, interpretability, and resource efficiency across diverse domains. In population-based optimization methods, the generation of diverse individuals holds utmost importance for adequately exploring the problem landscape, particularly in highly multi-modal multi-objective optimization problems. Our study reveals that, in line with findings from sev-eral prior research papers, commonly employed crossover and mutation operations lack the capability to generate high-quality diverse individuals and tend to become confined to limited areas around various local optima. This paper introduces an augmen-tation to the diversity of the population in the well-established multi-objective scheme of the genetic algorithm, NSGA-II. This enhancement is achieved through two key components: the genuine initialization method and the substitution of the worst individuals with new randomly generated individuals as a re-initialization approach in each generation. The proposed multi-objective feature selection method undergoes testing on twelve real-world classification problems, with the number of features ranging from 2,400 to nearly 50,000. The results demonstrate that replacing the last front of the population with an equivalent number of new random individuals generated using the genuine initialization method and featuring a limited number of features substantially improves the population's quality and, consequently, enhances the performance of the multi-objective algorithm.

Feature selection is a crucial step in the model-building pipeline in machine learning (ML) applications such as Electroencephalogram (EEG) signal processing, providing benefits on model performance and computational efficiency. EEG signals play a pivotal role in elucidating human nature through promising ML mechanisms. However, extracting a large number of features from the EEG signals can be a challenge of efficient EEG processing. A multi-objective feature selection strategy applied to the extracted features from EEG signals can simultaneously improve the accuracy of the process and reduce the number of features. However, the high dimensionality of the EEG feature vectors actually diminishes the exploration capabilities of multi-objective algorithms, impeding their real-world applicability. Hence, we propose a novel correlation-based multi-objective feature selection (COR-MFS) method, that aims to reduce dimensionality before applying the multi-objective algorithm. In the initial phase, a correlation-based dimension re-duction method is applied to filter the most relevant features with highest correlation with the class label. Subsequently, a multi-objective feature selection algorithm is applied to the shrunk search space enhancing optimization efficiency and facilitating the search process. We evaluated the proposed COR-MFS method using six large-scale EEG datasets and observed significant improvements compared to the stand-alone multi-objective feature selection. This underscores the effectiveness of our innovative framework in providing more accurate solutions with fewer number of features for extensive EEG-based classification tasks.

Feature selection is a demanding and costly endeavor within the realms of machine learning and data mining, targeting the elimination of irrelevant and redundant features. This endeavor significantly bolsters classification accuracy or other post processing components, such as search or clustering. In this context, the feature selection can be as a single-objective optimization task, with the primary objective of maximizing classification accuracy or multi-objective one while minimizing the number of selected features as well which can be tackled using population-based metaheuristic algorithms. Considering that feature selection (FS) inherently poses a binary problem, given that most metaheuristic algorithms operate in continuous domains (such as Real-coded GA, DE, PSO, and CMA-ES), transitioning them to binary search spaces necessitates substantial operator modifications. One of the crucial steps of the population-based algorithms is initialization of the population which significantly impacts both the convergence speed and the quality of the final solution. However, in most cases, especially with continuous algorithms, random initialization emerges as the predominant method for generating candidate solutions (initial population) which lacks diversity in the large-scale feature selection search spaces. In this paper, two novel population initialization methods within the continuous search space are introduced, followed by a comparative analysis against the traditional random initialization method. Experimental results conducted on ten datasets with 300 to 11,000 features demonstrate the effectiveness of population-level uniform initialization, surpassing the widely recognized individual-level uniform initialization method. The experiments are conducted using Differential Evolution as the single-objective algorithm and Generalized Differential Evolution3 as a multi-objective algorithm as our case studies. This study demonstrates the crucial role of initialization in the population-based optimization algorithms when they tackle binary problems.

Feature selection is a pivotal component of machine learning and data analysis, to optimize model performance by eliminating irrelevant and redundant features, to address the challenges associated with the “curse of dimensionality” and interpretability. In this context, we present feature selection as a multi-objective binary optimization task with the dual aim of maximizing classification accuracy while minimizing the number of selected features. In order to address this optimization challenge, we introduce the Multi-objective Binary Differential Evolution algorithm (MOBDE). It's worth noting that DE originally is an extremely powerful real-value coded algorithm, and to make it binary, set-based operators must replace vector-based operations. Optimization in binary space is deemed more suitable than real-value optimization because a many-to-one mapping can waste the efforts of the optimizer when solving a binary problem like feature selection. The algorithm leverages a partial opposition-based binary operator to generate diverse solutions to enhance its exploration within the search space. Additionally, it incorporates a majority voting mechanism as a local search strategy to bolster the algorithm's exploitation capabilities. Results from experimentation on eleven datasets underscore the efficiency of MOBDE, outperforming the widely recognized NSGA-II method in terms of the hypervolume (HV) performance metric and minimizing the number of selected features. The proposed algorithm and its experimental outcomes are comprehensively detailed and analyzed, offering valuable insights into its efficacy for feature selection tasks.

Deep neural networks (DNNs) have shown ex-ceptional performance in various domains, leading to their widespread adoption. However, the necessity to deploy DNNs on resource-constrained devices calls for improved model efficiency. Accordingly, DNN pruning has emerged as a critical technique in the field of machine learning, offering significant improve-ments in computational efficiency and model simplicity. This paper introduces an innovative post-training pruning approach for DNNs without any retraining, that utilizes multi-objective optimization to achieve substantial sparsity rates while preserving significant accuracy levels. The proposed method transforms the post-training weight pruning challenge into a two-variable, bi-objective optimization problem. The optimizer finds the optimal minimum and maximum threshold values through optimization, effectively converting the real-valued weights between these two thresholds to zero. The task and model-independency of the proposed framework make it applicable across various models, tasks, and datasets without constraints on the number of weights. The approach provides a decision-maker, where users can select the best strategy within resource constraints to achieve their desired accuracy. In order to assess our method, we evaluated the pruning of RESNET50 model on CIFAR10 and CIFAR100 benchmark datasets. In CIFAR10, by reducing 70% of the weights within the optimal threshold values, the network's accuracy only decreases by 0.1. Similarly, in CIFAR100, an appropriate weight range was selected, resulting in a 65 % reduction in weights while maintaining a negligible 0.1 decrease in network accuracy. This demonstrates the effectiveness of the optimization in achieving significant model size reduction without compromising performance on large DNNs.

The utilization of binary representation of the embeddings over real valued features represents a promising avenue, in terms of memory savings and faster operations for various machine learning models. In this research paper, we delve into the exploration of barcode representation for text embeddings derived from BERT, which is optimized using Co-ordinate Search algorithm. These binary embeddings present a compact representation of text, thereby mitigating memory and computational demands, which is especially advantageous in the context of resource-intensive large-scale text processing tasks. In our study, we introduce a novel optimal threshold technique, coupled with the Coordinate Search algorithm to transform continuous BERT embeddings into binary barcodes thereby enabling effective Natural Language Processing while sustaining computational efficiency. The optimal barcode representations have been applied in Natural Language Processing applications, showcasing its innovative potential in revolutionizing text representation. Through an extensive series of experiments on various NLP task encompassing diverse datasets, we comprehensively evaluate our approach, comparing it against a spectrum of thresholding techniques. The binary embeddings achieved by optimal thresholds outperform traditional binarization methods in terms of accuracy. The proposed method for generating a binary representations is versatile, being independent of the model, data and task, making it applicable across various machine learning applications.

Many real-world optimization problems can be modelled with several competing objectives. Most of the time, such optimization problems fall under the category of expensive problems. These are problems in which each fitness evaluation is time-consuming, for example, one fitness call could take hours or even days to compute. The time-consuming process of evaluating the function values or gradient of the objective functions may degrade the running speed of many optimization algorithms. The coordinate search (CS) approach is introduced as a single-objective gradient-free method for addressing large-scale, non-convex, and costly optimization problems. Due to the low computation and memory requirements of the CS algorithm, it can also be efficiently extended to address multi-objective optimization problems. The subject of this study is to develop a CS-based algorithm aimed at computationally expensive multi-objective optimization problems. In order to generate a set of non-dominated solutions, a population is created to apply the CS algorithm on each individual and finally reach an optimized interval. We demonstrate the efficacy of the proposed multi-objective CS method by comparing it with NSGA-II and MOEA/D as one of the well-known multi-objective algorithms on ZDT benchmark functions. Promising results are reported with the assumption of a limited number of fitness evaluations (NFF) which is desired during tackling complex and expensive optimization problems. Another major advantage of the proposed algorithm is that it provides regions of the Pareto front, that using sampling can generate as many Pareto front solutions as needed unlike other common optimization algorithms including NSGA-II.

Optimization problems widely arise in various science and engineering fields and can be computationally expensive in many real-world applications. Evaluation of the fitness function to assess a candidate solution is the main operation in all optimization procedures which can be heavily compute-intensive. Machine learning-based surrogate models can contribute to learning the specific pattern among the decision variables and objective values to consequently reduce the computation time of fitness evaluation. In this study, we have proposed a novel pairwise surrogate model to identify the superiority between candidate solutions in a pairwise comparison despite the fact that most of the surrogate models try to predict the exact fitness value. The proposed idea can significantly help the optimizer to reach better results in a shorter period of time. It seems comparing two candidate solutions for a greedy selection is much easier than approximating fitness values for both. We demonstrated Graph Neural Network (GNN) for this purpose to be trained on a limited number of pairwise ranks and then utilized to compare a pair of candidate solutions. In order to examine the efficacy of our model, we utilized different well-known single-objective optimization benchmarks in dimensions 10,20, and 30. Moreover, the results of the learning-based evaluation are compared with the results from the real fitness evaluation. The results, assessed in terms of the number of fitness calls and the best-found solution, showed that the proposed method is able to decrease the computing cost of fitness evaluation significantly while we achieve a comparable solution. Our model can be tested with any optimization algorithm which employs a comparison-based mechanism among its candidate solutions.

Generally speaking, labeled data is difficult and expensive to provide for applications in machine learning and data mining. One of the earliest approaches to tackle this problem is semi-supervised self-training to take advantages of labeled and unlabeled data to create pseudo-labeled data. However, reaching a high level of confidence to predict the unseen data by a classifier with a limited number of labeled samples is a challenging task. Generating pseudo-labeled data can collaboratively improve the self-training to increase its confidence for further predictions. This paper proposes a collaborative framework between augmentation and self-training to accurately train a model with very limited labeled data. Our framework includes two components where the first component is an unsupervised technique to augment labeled data and feed to the second component which is self-training. The first component uses a custom variational autoencoder architecture (VAE) to generate new samples by adding randomly generated noise to encoded latent representation. As a result, the augmentation component can generate the unique and unexplored images with respect to limited input data distribution. We evaluated the proposed framework on handwritten MNIST image dataset. The conducted experiment shows that the generative component can be helpful in overcoming the problem of inaccurate self-training prediction when sufficient labeled data is not accessible.

Feature selection is an expensive challenging task in machine learning and data mining aimed at removing irrelevant and redundant features. This contributes to an improvement in classification accuracy, as well as the budget and memory requirements for classification, or any other post-processing task conducted after feature selection. In this regard, we define feature selection as a multi-objective binary optimization task with the objectives of maximizing classification accuracy and minimizing the number of selected features. In order to select optimal features, we have proposed a binary Compact NSGA-II (CNSGA-II) algorithm. Compactness represents the population as a probability distribution to enhance evolutionary algorithms not only to be more memory-efficient but also to reduce the number of fitness evaluations. Instead of holding two populations during the optimization process, our proposed method uses several Probability Vectors (PVs) to generate new individuals. Each PV efficiently explores a region of the search space to find non-dominated solutions instead of generating candidate solutions from a small population as is the common approach in most evolutionary algorithms. To the best of our knowledge, this is the first compact multi-objective algorithm proposed for feature selection. The reported results for expensive optimization cases with a limited budget on five datasets show that the CNSGA-II performs more efficiently than the well-known NSGA-II method in terms of the hypervolume (HV) performance metric requiring less memory. The proposed method and experimental results are explained and analyzed in detail.

A supervised feature selection method selects an appropriate but concise set of features to differentiate classes, which is highly expensive for large-scale datasets. Therefore, feature selection should aim at both minimizing the number of selected features and maximizing the accuracy of classification, or any other task. However, this crucial task is computationally highly demanding on many real-world datasets and requires a very efficient algorithm to reach a set of optimal features with a limited number of fitness evaluations. For this purpose, we have proposed the binary multi-objective coordinate search (MOCS) algorithm to solve large-scale feature selection problems. To the best of our knowledge, the proposed algorithm in this paper is the first multi-objective coordinate search algorithm. In this method, we generate new individuals by flipping a variable of the candidate solutions on the Pareto front. This enables us to investigate the effectiveness of each feature in the corresponding subset. In fact, this strategy can play the role of crossover and mutation operators to generate distinct subsets of features. The reported results indicate the significant superiority of our method over NSGA-II, on five real-world large-scale datasets, particularly when the computing budget is limited. Moreover, this simple hyper-parameter-free algorithm can solve feature selection much faster and more efficiently than NSGA-II.

One of the main obstacles of adopting digital pathology is the challenge of efficient processing of hyperdimensional digitized biopsy samples, called whole slide images (WSIs). Exploiting deep learning and introducing compact WSI representations are urgently needed to accelerate image analysis and facilitate the visualization and interpretability of pathology results in a postpandemic world. In this article, we introduce a new evolutionary approach for WSI representation based on large-scale multiobjective optimization (LSMOP) of deep embeddings. We start with patch-based sampling to feed KimiaNet, a histopathology-specialized deep network, and to extract a multitude of feature vectors. Coarse multiobjective feature selection uses the reduced search space strategy guided by the classification accuracy and the number of features. In the second stage, the frequent features histogram (FFH), a novel WSI representation, is constructed by multiple runs of coarse LSMOP. Fine evolutionary feature selection is then applied to find a compact (short-length) feature vector based on the FFH and contributes to a more robust deep-learning approach to digital pathology supported by the stochastic power of evolutionary algorithms. We validate the proposed schemes using The Cancer Genome Atlas (TCGA) images in terms of WSI representation, classification accuracy, and feature quality. Furthermore, a novel decision space for multicriteria decision making in the LSMOP field is introduced. Finally, a patch-level visualization approach is proposed to increase the interpretability of deep features. The proposed evolutionary algorithm finds a very compact feature vector to represent a WSI (almost 14000 times smaller than the original feature vectors) with 8% higher accuracy compared to the codes provided by the state-of-the-art methods

In recent years, center-based sampling has demonstrated impressive results to enhance the efficiency and effectiveness of meta-heuristic algorithms. The strategy of the center-based sampling can be utilized at either or both the operation and/or population level. Despite the overall efficiency of the center-based sampling in population-based algorithms, utilization at operation level requires customizing the strategy for a specific algorithm which degrades the scheme's generalization. In this paper, we have proposed a population-level center-based sampling method which is operation independent and correspondingly can be embedded in any population-based optimization algorithm. In this study, we applied the proposed scheme for Differential Evolution (DE) algorithm to enhance the exploration and exploitation capabilities of the algorithm. We cluster candidate solutions and inject the centroid-based samples into the population to increase the overall quality of the population and thus decrease the risk of premature convergence and stagnation. By a high chance, the center-based samples are effectively generated in the promising regions of the search space. The proposed method has been benchmarked by employing CEC-2017 benchmark test suite on dimensions 30, 50, and 100. The results clearly indicate the superiority of the proposed scheme, and a detailed results analysis is provided.

Deep neural networks (DNNs) show impressive performance for hyperspectral image (HSI) classification when abundant labeled samples are available. The problem is that HSI sample annotation is extremely costly and the budget for this task is usually limited. To reduce the reliance on labeled samples, deep semi-supervised learning (SSL), which jointly learns from labeled and unlabeled samples, has been introduced in the literature. However, learning robust and discriminative features from unlabeled data is a challenging task due to various noise effects and ambiguity of unlabeled samples. As a result, recent advances are constrained, mainly in the pre-training or warm-up stage. In this paper, we propose a deep probabilistic framework to generate reliable pseudo labels to explicitly learn discriminative features from unlabeled samples. The generated pseudo labels of our proposed framework can be fed to various DNNs to improve their generalization capacity. Our proposed framework takes only 10 labeled samples per class to represent the label set as an uncertainty-aware distribution in the latent space. The pseudo labels are then generated for those unlabeled samples whose feature values match the distribution with high probability. By performing extensive experiments on four publicly available datasets, we show that our framework can generate reliable pseudo labels to significantly improve the generalization capacity of several state-of-the-art DNNs. In addition, we introduce a new DNN for HSI classification that demonstrates outstanding accuracy results in comparison with its rivals.

Differential Evolution (DE) is one of the well-established population-based optimization algorithms which has received a lot of attention regarding its potential to solve complex optimization problems. However, DE is capable to explore a huge search space in its early run phase, called exploration phase, its weakness in exploitation avoids local refinement of the promising shrunk region. Therefore, employing a local search can be an efficient strategy to improve the search performance of DE via accelerating of fine tuning phase. This paper purposes an effective Memetic DE algorithm using a well-known single-solution-based optimization method, i.e., Coordinate Descent (CD) algorithm. Local coordinate search is applied on the promising region resulted by top ranked individuals selected from the final population of DE. The proposed method updates the value of each coordinate iteratively by evaluating the sampled points from the local region to improve the resulted candidate solution. Since coordinate search algorithm shrinks the region rapidly, it requires a very small portion of the computational budget to find the optimal coordinates' value. In order to evaluate the proposed Memetic DE, several experiment series are conducted on functions of CEC-2017 benchmark for different number of dimensions (i.e., D=30, 50, and 100). Results clearly indicate that the utilized local coordinate search improves the quality of resulted solution by DE significantly using a very low computational budget, i.e., 20×D.

Population-based algorithms are a well-established category of metaheuristic optimization algorithms in which individuals collaborate with each other to find the optimal solution in a search space. During the search process, each individual provides a partial intelligence which can assist the population movement toward promising regions. In this paper, a dimension-wise strategy is proposed to collect the intelligence of whole population to generate a new trial candidate solution. For new individual, the value of each variable is calculated using the votes of a more-crowded cluster of individuals obtained on each dimension (one-dimensional clustering). Accordingly, a group of candidate solutions in the population collaborate to determine a variable value of new individual. Utilizing this strategy, collective intelligence (CI) aims the algorithm to find better candidate solutions. Since the proposed method keeps untouched all other parts of the algorithm, it can be used with any population-based algorithm. This paper presents the modification of two well-known population-based algorithms based on the proposed strategy in utilizing Collective Intelligence (CI), Differential Evolution (CIDE) and Particle Swarm optimization (CIPSO). In conducted experiments, two proposed algorithms are compared with classical version of DE and PSO on 30 functions of CEC2017 benchmark. The results indicate that the proposed method generates an individual with better objective function value than many of the individuals in the population which leads totally better results in overall.

Solving real-world complex optimization problems using simple metaheuristic algorithms is a challenging but attractive task. Making matrix multiplication efficient is one of the interesting problems. This time-consuming algebric operation is required in many applications in science and engineering, thus reducing its complexity targets more efficient computation. In fact, many of practical and theatrical complicated calculations can be modeled efficiently as matrix-based operations, therefore matrix multiplication is computationally expansive operator among all others. In this paper, a simple metaheuristic method based on Micro Genetic Algorithm is proposed to find Strassen's equivalent solutions which is an algebraic method to compute the product of two matrices with minimal number of multiplications. Since there are numerous optimal solutions, the modeled problem is a large-scale and highly multi-modal optimization problem. The proposed method could find more than 160, 000 valid solutions with same complexity as Strassen's in a large-scale search space. Among all discovered solutions found using the proposed method, there are 701 distinct solutions which is the maximum number of discovered Strassen's equivalent solutions to the best of our knowledge. The proposed algorithm is simple but very efficient to find more and more solutions, in fact, that is a great demonstration of “evolution in action” to tackle real-world complex problems like the current one, which just one set of its equations has been discovered by the Germen mathematicians and has remained mysterious for more than 50 years.

Crossover is the main genetic operator which influences the power of evolutionary algorithms. Among a variety of crossover operators, there has been a growing interest in multi-parent crossover operators in evolutionary computation. The main motivation of those schemes is establishing comprehensive collective collaboration of more than two chromosomes in the population to generate a new offspring. In this paper, a novel all-parent crossover operator called collective crossover for genetic algorithm is proposed. In this method, all individuals in the current population are involved in recombination part and one offspring is generated. The contribution of each individuals is defined based on its quality in terms of fitness value. The performance of the collective crossover operator is tested on CEC2017 benchmark functions. The results revealed that the proposed crossover operator performs better when compared to well-known two-parent crossover operators including one-point and two-point crossovers. In addition, the differences between collective crossover and the other crossover operators are statistically significant for the most cases.

In machine learning, multi-label classification aims to assign labels of instances in a dataset which are associated to more than one class label. Feature selection as an important task in predictive model construction improves the performance of multi-label classification. Since feature selection task can be interpreted as optimizing multiple objectives in a massive search space, multi-objective evolutionary techniques can be applied to tackle this family of problems. In this paper, a binary multi-objective feature selection is proposed for multi-label data with considering number of features and classification accuracy as objectives. A binary differential evolution is proposed based on opposition-based learning concept and partially voting between two candidate solutions to decide about absence or presence of a feature in third randomly selected solution. Because feature selection is basically a binary optimization problem, proposing a binary operator improves the effectiveness of search process in evolutionary algorithms. The proposed operator is utilized in third version of Generalized Differential Evolution (GDE3) which is a multi-objective optimization algorithm to select best subset of multi-label features with minimum number of features. A benchmarking is conducted on eight multi-label datasets in terms of several multi-objective assessment metrics including the Hypervolume indicator, Pure Diversity, and Set-coverage. Experimental results show significant improvements for proposed method in comparison with the state-of-the-art multi-objective feature selection methods for multi-label classification, which are namely NSGA-II and PSO based approaches.

Scientists' publication impacts ranking is an important topic in scientometrics which is performed based on various proposed criteria. One of the well-known indicators is h-index which evaluates researchers achievements based on number of citations. The h-index has utilized in many research data sources because of its appropriate properties, but similar to other assessment indicators, it has own disadvantages. hindex cannot give a fair comparison between junior and senior researches. There are two reasons for this unfair comparison: (1) h-index depends on the research period of scholars and (2) the number of received citations can be increased by time, even if researcher doesn't publish new papers, the h-index increases. Consequently, in addition to h-index, the number of the years of academic research (called the research period) is preferable to be considered as an independent indicator, which makes us able to have a more fair evaluation. So these two objectives, maximizing h-index and minimizing research period, can be considered as a multi-criteria comparison task to assess researchers. In this paper, we propose a strategy based on Pareto dominance ranking which uses dominance concept to obtain an order for researchers. In order to complete ranking between scientists in the same rank, a multi-criteria decision making measure called VIKOR is utilized. Therefore, a total ranking measure (P-V index) is obtained using Perto front concept and VIKOR measure. The proposed method is applied on 235 researchers who are conducting research on Evolutionary Computation (EC) topic. The h-index value and the research period of scholars are collected via Google Scholar service. P-V index obtains 26 Pareto ranks for all researchers and places six EC scientists on the first Pareto front.

For several years, the Differential Evolution (DE) algorithm has been an effective method for solving complex real-world optimization problems. Due to its success and popularity, there are several multi-objective optimization algorithms proposed based on DE. However, when DE comes to solving large-scale problems its performance deteriorates. Several recent studies clearly confirm that utilizing center-based sampling method can increase the probability of the closeness of initialized population individuals to the solutions in black-box problems. In this paper, we propose center-based mutation for Third Generalized Differential Evolution (CGDE3) algorithm in order to solve large-scale multi-objective optimization problems; in fact, this time center-based sampling scheme is employed during the optimization process not just during the population initialization phase. For its mutation scheme, the CGDE3 algorithm utilizes five randomly selected candidate solutions from its current population to generate a new trial vector. The proposed method enhances the GDE3 algorithm by improving its exploration ability using extra center-based sampling during the evolution process. This algorithm is tested on benchmarks of CEC 2017 competition on evolutionary multi-objective optimization with dimensions of 100, 500 and 1000. Experimental results confirm that CGDE3 outperforms GDE3 over all three studied large-scale dimensions.