Solving dynamic multi-objective optimization problems via support vector machine

Dynamic Multi-objective Optimization Problems (DMOPs) refer to optimization problems that objective functions will change with time. Solving DMOPs implies that the Pareto Optimal Set (POS) at different moments can be accurately found, and this is a very difficult job due to the dynamics of the optimization problems. The POS that have been obtained in the past can help us to find the POS of the next time more quickly and accurately. Therefore, in this paper we present a Support Vector Machine (SVM) based Dynamic Multi-Objective Evolutionary optimization Algorithm, called SVM-DMOEA. The algorithm uses the POS that has been obtained to train a SVM and then take the trained SVM to classify the solutions of the dynamic optimization problem at the next moment, and thus it is able to generate an initial population which consists of different individuals recognized by the trained SVM. The initial populuation can be fed into any population based optimization algorithm, e.g., the Nondominated Sorting Genetic Algorithm II (NSGA-II), to get the POS at that moment. The experimental results show the validity of our proposed approach.

[1]  Changhe Li,et al.  Fast Multi-Swarm Optimization for Dynamic Optimization Problems , 2008, 2008 Fourth International Conference on Natural Computation.

[2]  Kay Chen Tan,et al.  Evolutionary Dynamic Multiobjective Optimization Via Kalman Filter Prediction , 2016, IEEE Transactions on Cybernetics.

[3]  Lamjed Ben Said,et al.  Dynamic Multi-objective Optimization Using Evolutionary Algorithms: A Survey , 2017, Recent Advances in Evolutionary Multi-objective Optimization.

[4]  Fan Zhang,et al.  Fuzzy neural network based dynamic path planning , 2012, 2012 International Conference on Machine Learning and Cybernetics.

[5]  Kalyanmoy Deb,et al.  Dynamic Multi-objective Optimization and Decision-Making Using Modified NSGA-II: A Case Study on Hydro-thermal Power Scheduling , 2007, EMO.

[6]  T. Jansen,et al.  Populations can be Essential in Dynamic Optimisation , 2015, GECCO.

[7]  Jürgen Branke,et al.  Memory enhanced evolutionary algorithms for changing optimization problems , 1999, Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat. No. 99TH8406).

[8]  Min Jiang,et al.  Integration of Global and Local Metrics for Domain Adaptation Learning Via Dimensionality Reduction , 2017, IEEE Transactions on Cybernetics.

[9]  Duc-Cuong Dang,et al.  Populations Can Be Essential in Tracking Dynamic Optima , 2016, Algorithmica.

[10]  Min Jiang,et al.  Embodied concept formation and reasoning via neural-symbolic integration , 2010, Neurocomputing.

[11]  Johan A. K. Suykens,et al.  Least Squares Support Vector Machine Classifiers , 1999, Neural Processing Letters.

[12]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[13]  Carlos A. Coello Coello,et al.  Improving PSO-Based Multi-objective Optimization Using Crowding, Mutation and epsilon-Dominance , 2005, EMO.

[14]  Kay Chen Tan,et al.  Solving the IEEE CEC 2015 Dynamic Benchmark Problems Using Kalman Filter Based Dynamic Multiobjective Evolutionary Algorithm , 2016 .

[15]  Lamjed Ben Said,et al.  A dynamic multi-objective evolutionary algorithm using a change severity-based adaptive population management strategy , 2015, Soft Computing.

[16]  Gary G. Yen,et al.  Dynamic Evolutionary Algorithm With Variable Relocation , 2009, IEEE Transactions on Evolutionary Computation.

[17]  A. Engelbrecht,et al.  Benchmark Functions for CEC 2015 Special Session and Competition on Dynamic Multi-objective Optimization , 2015 .

[18]  Shengxiang Yang,et al.  Evolutionary dynamic optimization: A survey of the state of the art , 2012, Swarm Evol. Comput..

[19]  Bernhard E. Boser,et al.  A training algorithm for optimal margin classifiers , 1992, COLT '92.

[20]  Claudio Rossi,et al.  Tracking Moving Optima Using Kalman-Based Predictions , 2008, Evolutionary Computation.

[21]  Kalyanmoy Deb,et al.  Dynamic multiobjective optimization problems: test cases, approximations, and applications , 2004, IEEE Transactions on Evolutionary Computation.

[22]  C. Coello,et al.  Improving PSO-based Multi-Objective Optimization using Crowding , Mutation and �-Dominance , 2005 .

[23]  Gary G. Yen,et al.  Transfer Learning-Based Dynamic Multiobjective Optimization Algorithms , 2016, IEEE Transactions on Evolutionary Computation.

[24]  Qingfu Zhang,et al.  A Population Prediction Strategy for Evolutionary Dynamic Multiobjective Optimization , 2014, IEEE Transactions on Cybernetics.

[25]  Concha Bielza,et al.  Multiobjective Estimation of Distribution Algorithm Based on Joint Modeling of Objectives and Variables , 2014, IEEE Transactions on Evolutionary Computation.

[26]  Carlos Cruz,et al.  Optimization in dynamic environments: a survey on problems, methods and measures , 2011, Soft Comput..

[27]  Ben Goertzel,et al.  Improving machine vision via incorporating expectation-maximization into Deep Spatio-Temporal learning , 2014, 2014 International Joint Conference on Neural Networks (IJCNN).

[28]  Peter A. N. Bosman.  Learning and Anticipation in Online Dynamic Optimization , 2007, Evolutionary Computation in Dynamic and Uncertain Environments.

[29]  Helen G. Cobb,et al.  An Investigation into the Use of Hypermutation as an Adaptive Operator in Genetic Algorithms Having Continuous, Time-Dependent Nonstationary Environments , 1990 .