Showing papers in "Applied Soft Computing in 2018"

TL;DR: A review on deep learning methods for semantic segmentation applied to various application areas and points out a set of promising future works to help researchers decide which are the ones that best suit their needs and goals.

TL;DR: A new wrapper feature selection approach is proposed based on Whale Optimization Algorithm based on the influence of using the Tournament and Roulette Wheel selection mechanisms instead of using a random operator in the searching process to search the optimal feature subsets for classification purposes.

TL;DR: A two phase hybrid model for cancer classification is being proposed, integrating Correlation-based Feature Selection (CFS) with improved-Binary Particle Swarm Optimization (iBPSO), which achieves up to 100% classification accuracy for seven out of eleven datasets with a very small sized prognostic gene subset.

TL;DR: The solution results quality of this study show that the proposed HFPSO algorithm provides fast and reliable optimization solutions and outperforms others in unimodal, simple multi-modal, hybrid, and composition categories of computationally expensive numerical functions.

TL;DR: A novel algorithmic trading model CNN-TA is proposed using a 2-D convolutional neural network based on image processing properties that provides better results for stocks and ETFs over a long out-of-sample period.

TL;DR: The farmland fertility in problems with smaller dimensions problems has been able to act as a strong metaheuristic algorithm and it has optimized problems nicely and the effectiveness of other algorithms decreases significantly with number of dimensions and the farmland fertility obtains better results than other algorithms.

TL;DR: A fog-based attack detection framework that relies on the fog computing paradigm and a newly proposed ELM-based Semi-supervised Fuzzy C-Means method, demonstrating that the proposed framework achieved better performance than the centralized attack Detection framework.

TL;DR: The results of sensitivity and comparative analyses show that the proposed integrated fuzzy MCDM approach for FMEA is valid and can provide valuable and effective information in assisting risk management decision-making.

TL;DR: Results prove that the proposed Hybrid SCA-DE-based tracker can robustly track an arbitrary target in various challenging conditions than the other trackers and is very competitive compared to the state-of-the-art metaheuristic algorithms.

TL;DR: A novel approach to feature selection in credit scoring applications is proposed, called Information Gain Directed Feature Selection algorithm (IGDFS), which performs the ranking of features based on information gain, propagates the top m features through the GA wrapper (GAW) algorithm using three classical machine learning algorithms of KNN, Naive Bayes and Support Vector Machine for credit scoring.

TL;DR: The IABC algorithm is proposed to improve the evacuation efficiency and provide support for building design and evacuation management by employing the strategies of grouping and exit selection and uses the evacuation time of the individuals as the evaluation metric.

TL;DR: In this study, a multi-criteria decision-making framework is constructed for risk evaluation of construction project with picture fuzzy information and an integrated picture fuzzy normalized projection-based VIKOR method is constructed by integrate the PFNP model and VIKor method under picture fuzzy environment.

TL;DR: Using combination of the SCA and Levy flight in the PSOSCALF algorithm, the exploration capability of the original PSO algorithm is enhanced and also, being trapped in the local minimum is prevented.

TL;DR: It is verified that utilizing half of the salps as leaders of the chain can significantly improve the performance of SSA in terms of accuracy metric and dynamically tuning the single parameter of algorithm enable it to more effectively explore the search space in dealing with different feature selection datasets.

TL;DR: Results show that the VPL algorithm possesses a strong capability to produce superior performance over the other well-known metaheuristic algorithms and is effectively applicable to solve problems with complex search space.

TL;DR: A neuro-heuristic scheme is design to solve nonlinear singular second order system based on Thomas-Fermi equation using the strength of universal approximation capabilities of feedforward artificial neural networks supported with optimization power of genetic algorithms and sequential quadratic programming.

TL;DR: This work proposes a novel feature selection approach designed to deal with two major issues in machine learning, namely class-imbalance and high dimensionality, and achieves the highest average predictive performance with the approach compared with the most well-known feature selection strategies.

TL;DR: The credibility of the IR-AHP-MABAC model was demonstrated by comparing the results of different multi-criteria techniques and analyzing viability, and shown that the new approach to dealing with imprecision yields credible, reputable ranks.

TL;DR: The ANNs and its particular structure can be successfully utilized and modeled as metaheuristic optimization method for handling optimization problems and an iterative convergence of NNA is proved theoretically.

TL;DR: A new model that uses multiple-criteria decision-making in combination with grey theory for FMEA is proposed, which preserves the information of prioritized failure modes through interval analysis and can provide an alternative risk priority solution for product development.

TL;DR: A multi-sensor information fusion system for online RUL prediction of machining tools is proposed and the system includes sensor signal preprocessing based on ensemble empirical mode decomposition method, statistics feature extraction based on time domain and frequency domain analysis, optimum feature selection based on Pearson correlation coefficient and feature fusion based on adaptive network based fuzzy inference system.

TL;DR: The proposed MCSA develops the search capability of crows in the original CSA and introduces a new way by which a destination is selected by a crow to follow and is applied on five different well-known test systems to indicate the applicability of MCSA in the ELD problem.

TL;DR: An overview of the application of computational intelligence technologies in optical remote sensing image processing, including: 1) feature representation and selection; 2) classification and clustering; and 3) change detection are provided.

TL;DR: A novel hybrid algorithm based on BBO and Grey Wolf Optimizer, named HBBOG, is presented, which can effectively maximize the two algorithms’ advantages and overall balance exploration and exploitation and obtain strong universal applicability.

TL;DR: The soft computing model is presented as a simple formula and excellent agreement is obtained representing a high degree of reliability for the proposed model.

TL;DR: A novel hybrid forecasting model based on an improved grey forecasting mode optimized by multi-objective ant lion optimization algorithm is successfully developed, which can not only be utilized to dynamic choose the best input training sets, but also obtain satisfactory forecasting results with high accuracy and strong ability.

TL;DR: A novel approach is proposed, which utilizes the mechanism of Reinforcement Learning (RL) to obtain high accurate QoE in resource allocations in IoT and utilizes the satisfactory level of Quality of Experience (QoE) to achieve intelligent content-centric services.

TL;DR: This paper investigates the ELECTre II method in the HFLTS environment and proposes two new approaches named the score-deviation-based ELECTRE II method and the positive and negative ideal hesitant fuzzy linguistic elements based ELECTRE I method.

TL;DR: The computational experiments show that the proposed Hybrid Ant Colony algorithm provides better results relative to the other algorithms, compared to the Adaptive Learning Approach and Genetic Heuristic algorithm.

TL;DR: Examination of experimentally how granularity level affects both the classification accuracy and the size of feature subset for feature selection shows that the approaches are efficient and can provide higher classification accuracy using granular information.