查看更多>>摘要：The speed by which the COVID-19 pandemic spread throughout the world makes the emergency services unprepared to answer all the patients’ requests. The Tunisian ministry of health established a protocol planning the sample collection from the patients at their location. A triage score is first assigned to each patient according to the symptoms he is showing, and his health conditions. Then, given the limited number of the available ambulances in each area, the location of the patients and the capacity of the nearby hospitals for receiving the testing samples, an ambulance scheduling and routing plan needs to be established so that specimens can be transferred to hospitals in short time. In this paper, we propose to model this problem as a Multi-Origin–Destination Team Orienteering Problem (MODTOP). The objective is to find the optimal one day tour plan for the available ambulances that maximizes the collected scores of visited patients while respecting duration and capacity constraints. To solve this NP-hard problem, two highly effective approaches are proposed which are Hybrid Genetic Algorithm (HGA) and Memetic Algorithm (MA). The HGA combines (i) a k-means construction method for initial population generation and (ii) a one point crossover operator for solution recombination. The MA is an improvement of HGA that integrates an effective local search based on three different neighborhood structures. Computational experiments, supported by a statistical analysis on benchmark data sets, illustrate the efficiency of the proposed approaches. HGA and MA reached the best known solutions in 54.7% and 73.5% of instances, respectively. Likewise, MA reached a relative error of 0.0675% and performed better than four existing approaches. Real-case instances derived from the city of Tunis were also solved and compared with the results of an exact solver Cplex to validate the effectiveness of our algorithm.
查看更多>>摘要：Classifier design is one of the most significant fields in pattern recognition. Most classifiers are measured by classification accuracy, which assumes that all the misclassification cost are the same. In the real world, different misclassifications usually bring different losses. Based on this fact, cost-sensitive learning is becoming a hot research area in pattern recognition. However, in cost-sensitive learning, examples costs are often difficult to achieve and usually decided by the authors experience. Hence, combining the cost-sensitive learning and matrixized learning thoughts, we propose a two-class cost-sensitive matrixized classification model based on information entropy called CsMatMHKS in this paper. The proposed CsMatMHKS introduces information entropy which can reveal the uncertainty of one sample into matrixized learning framework to decrease the total misclassification cost. The experimental results on the UCI datasets and image datasets indicate that the CsMatMHKS not only reduces the sum of classification costs but also has comparable classification accuracy.
查看更多>>摘要：This article introduces a version of the Self-Organizing Migrating Algorithm with a narrowing search space strategy named iSOMA. Compared to the previous two versions, SOMA T3A and Pareto that ranked 3rd and 5th respectively in the IEEE CEC (Congress on Evolutionary Computation) 2019 competition, the iSOMA is equipped with more advanced features with notable improvements including applying jumps in the order, immediate update, narrowing the search space instead of searching on the intersecting edges of hyperplanes, and the partial replacement of individuals in the population when the global best improved no further. Moreover, the proposed algorithm is organized into processes named initialization, self-organizing, migrating, and replacement. We tested the performance of this new version by using three benchmark test suites of IEEE CEC 2013, 2015, and 2017, which, together contain a total of 73 functions. Not only is it superior in performance to other SOMAs, but iSOMA also yields promising results against the representatives of well-known algorithmic families such as Differential Evolution and Particle Swarm Optimization. Moreover, we demonstrate the application of iSOMA for path planning of a drone, while avoiding static obstacles and catching the target.
查看更多>>摘要：Toward higher solution efficiency and faster computation, this paper addresses the minimum weighted vertex cover (MWVC) problem by introducing game theory into iterated optimization and proposing a population-based game-theoretic optimizer (PGTO). A group of candidate solutions are iterated through the specially designed procedures of swarm evolution (SE), learning in games (LIG), and local search (LS) successively. Within the framework of potential game theory, we prove that LIG computes in finite time Nash equilibria that represent vertex cover solutions where no redundant nodes exist. Moreover, theoretical analysis is presented that by exchanging the actions of certain neighbours, LS is capable of generating better results upon the input Nash equilibrium. Through intensive numerical experiments, we show that while enlarging the population size could boost the global objective, a mutation probability between 0.05 and 0.2 is more likely to provide the best performance. Comparison experiments against the state of the art demonstrate the superiority of the presented methodology, both in terms of solution quality and computation speed.
查看更多>>摘要：This paper presents the application of a hybrid model to predict the mid-price trend of assets in the Brazilian Stock Exchange (B3) from the market by order data. A Convolution Neural Network (CNN) is applied to extract spatial features from an order book aggregated by price and then a decision tree-based algorithm (CatBoost) combines these CNN features with events provided by Times and Trades information (TTinfo) to have the final prediction. Differently from most stock exchanges, TTinfo from B3 includes to which broker an order belongs, and in this work its impact in the final prediction is analysed as well. The proposed solution innovates by joining CNN with CatBoost, improving accuracy by 8% when compared to a common CNN, where 5% is only due to the adoption of CatBoost and another 3% is due to the combination of features from CNN with TTinfo. In addition, for training update, only CatBoost needs to be retrained, allowing learning transfer for the CNN, which reduces the overall updating time in at least one order of magnitude.
查看更多>>摘要：Supply Chain Management is frequently regarded as a distributed system whose productivity is mainly influenced by healthy interaction and cooperation among members. Examples of such inefficiencies can be seen in “Demand amplification” referring to asymmetric increase of demands among supply echelons due to both operational and behavioral causes. This paper addresses defective loops of non-cooperation as well as absence of information to manage the dynamics of demand amplification in a four-echelon supply chain; a new agent-based structure is suggested to facilitate the cooperation and coordination among major components and provide a structured context for interactive information sharing. Adequate motivation to share the required information is derived from an automated negotiation between retailer and manufacturer echelons in a four-echelon serial supply model. In doing so, the retailer agent would be encouraged to share customer demand information using Token-Based ordering policy, through a Reverse Ultimatum Game negotiation module. A novel fuzzy approach is suggested in order to cope with ambiguities involved in this negotiation; Numerical experiments prove that the proposed fuzzy negotiation mechanism can warrant the agreement among negotiating parties in nearly half the number of RUG iterations with 30% agreement share. This success in bringing the negotiation parties to an agreement results in the bullwhip effect decreasing by 30% in a four-echelon agent-based supply system. The proposed agent-based supply model which is designed to facilitate this cooperative decision-making process includes some unique innovative features: the knowledge base of the proposed system is able to retrieve and reuse the previous negotiation outcomes, through a gradual learning module. A combination of Case-Based Reasoning and Rule-Based inference mechanisms are applied to facilitate this, so prior cases will be stored in a Frame-Based structure. Eventually, the integrity of the proposed agent-based system is examined through combining Top-Down and Subsets integration approaches and some numerical experiments are provided to confirm the efficiency of the proposed agent-based structure in bullwhip effect management.
查看更多>>摘要：B-cell epitope prediction research has received growing interest since the development of the first method. B-cell epitope identification with the aid of an accurate prediction method is one of the most important steps in epitope-based vaccine development, immunodiagnostic testing, antibody production, disease diagnosis, and treatment. Nevertheless, using experimental methods in epitope mapping is very time-consuming, costly, and labor-intensive. Therefore, although successful predictions with in silico methods are very important in epitope prediction, there are limited studies in this area. The aim of this study is to propose a new approach for successfully predicting B-cell epitopes for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). In this study, the SARS-CoV B-cell epitope prediction performances of different fuzzy learning classification models genetic cooperative competitive learning (GCCL), fuzzy genetics-based machine learning (GBML), Chi's method (CHI), Ishibuchi's method with weight factor (W), structural learning algorithm on vague environment (SLAVE) and the state-of-the-art ensemble fuzzy classification model were compared. The obtained results showed that the proposed ensemble approach has the lowest error in SARS-CoV B-cell epitope estimation compared to the base fuzzy learners (average error rates; ensemble fuzzy=8.33, GCCL=30.42, GBML=23.82, CHI=29.17, W=46.25, and SLAVE=20.42). SARS-CoV and SARS-CoV-2 have high genome similarities. Therefore, the most successful method determined for SARS-CoV B-cell epitope prediction was used in SARS-CoV-2 cell epitope prediction. Finally, the eventual B-cell epitope prediction results obtained for SARS-CoV-2 with the ensemble fuzzy classification model were compared with the epitope sequences predicted by the BepiPred server and immunoinformatics studies in the literature for the same protein sequences according to VaxiJen 2.0 scores. We hope that the developed epitope prediction method will help design effective vaccines and drugs against future outbreaks of the coronavirus family, especially SARS-CoV-2 and its possible mutations.
查看更多>>摘要：Several DataGlove wearable electronic devices based on real-time Sign Language Recognition Systems (SLRSs) have been developed recently to assist the deaf and dumb community in translating hand gestures to their spoken language equivalents. Multidimensional evaluation and benchmarking of these systems are critical for determining the most desirable approach for meeting all essential requirements. However, this process is considered a Multicriteria Decision-Making (MCDM) problem due to the presence of several issues, including multiple evaluation criteria, criteria importance and criteria confliction. Hence, the MCDM approach is required to solve these issues. In this study, a new extension of the Fuzzy Decision by Opinion Score Method (FDOSM) for evaluating and benchmarking SLRSs is developed under an Interval-Valued Pythagorean Fuzzy Set (IVPFS) named IVP-FDOSM. Fundamentally, the methodology includes two phases. In the first phase, a decision matrix is formulated on the basis of identified ‘multidimensional criteria of hand gesture recognition and sensor glove perspectives’ and ‘real-time SLRSs’. The second phase introduces the development of IVP-FDOSM in two stages. The decision matrix is transformed into an opinion matrix in the first stage (data transformation unit). Meanwhile, in the second stage (data-processing unit), the opinion matrix is converted into fuzzy opinion decision matrices with the assistance of three experts by transforming the opinion matrix's linguistic terms to Interval-Value Pythagorean Fuzzy Numbers (IVPFNs). Results indicate the following: (1) individual benchmarking results of real-time SLRS showed high variation (90%) based on the preference of each Decision Maker (DM), with only 10% of preferences being identical. (2) The results of group benchmarking reveal that the 10th real-time SLRS was the optimal one and the 6th was the worst. In addition, the rates of ranking match between the group benchmarking and each DM (first DM, second DM and third DM) were 23%, 37% and 43%, respectively. (3) For the results evaluation, the statistical test-based objective assessment shows that the first group received the lowest mean value (2.80333), while the third group received the highest mean (3.8). These values indicate that the group benchmarked systems resulting from IVP-FDOSM are undergoing a systematic ranking. Furthermore, the comparative analysis reveals that IVP-FDOSM is superior to IVP-TOPSIS and IVP-AHP in terms of ranking and weighting.
查看更多>>摘要：Non-linear regression is the primary tool for estimating kinetic models of chemical reactions. The default approach of minimizing the sum of squared residuals tends to underperform in the presence of systematic errors, non-normal distribution of residuals or identifiability issues such as a high correlation between parameters. Therefore, we argue for a careful choice of the fit criteria and propose new, concave loss functions. Together with regularization, they form a robust objective for the regression procedure. Discussion of the rationale behind the proposed approach and its effects is illustrated by laboratory data on the transesterification of palm oil. A dedicated simulation study complements qualitative examples. All of the top-performing methods use regularization. Concave loss functions were among the best in 6–7 out of 8 test cases, compared to 2–3 for the classical square loss confirming both statistical and practical usefulness of the novel fit criteria. This result holds for a variety of modern optimizers. In 76% of our simulations, we obtained results not significantly worse than the best, whereas methods currently used in the literature provide 38% for the relative and 0% for the square loss.
查看更多>>摘要：The world is currently experiencing an ongoing pandemic of an infectious disease named coronavirus disease 2019 (i.e., COVID-19), which is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Computed Tomography (CT) plays an important role in assessing the severity of the infection and can also be used to identify those symptomatic and asymptomatic COVID-19 carriers. With a surge of the cumulative number of COVID-19 patients, radiologists are increasingly stressed to examine the CT scans manually. Therefore, an automated 3D CT scan recognition tool is highly in demand since the manual analysis is time-consuming for radiologists and their fatigue can cause possible misjudgment. However, due to various technical specifications of CT scanners located in different hospitals, the appearance of CT images can be significantly different leading to the failure of many automated image recognition approaches. The multi-domain shift problem for the multi-center and multi-scanner studies is therefore nontrivial that is also crucial for a dependable recognition and critical for reproducible and objective diagnosis and prognosis. In this paper, we proposed a COVID-19 CT scan recognition model namely coronavirus information fusion and diagnosis network (CIFD-Net) that can efficiently handle the multi-domain shift problem via a new robust weakly supervised learning paradigm. Our model can resolve the problem of different appearance in CT scan images reliably and efficiently while attaining higher accuracy compared to other state-of-the-art methods.
NSTL
Elsevier