Research On Rough Set Attribute Reduction Method Using Ant Colony Algorithm

With the proliferation of high-dimensional data,feature selection has become an indispensable task in the learning process.Rough set(Rough Set)theory is an effective mathematical tool to deal with imprecise,incomplete and uncertain.The problem of attribute reduction in rough set theory has been recognized as an important method of feature selection.As we all know,attribute reduction is an NP-hard problem.Since some basic attribute reduction methods can not find the optimal solution,many of the research work has been transferred to modern heuristic algorithms such as genetic algorithms,ant colony optimization(ACO),simulated annealing(SA),tabu search(TS),and particle swarm optimization(PSO),which can get higher quality solutions.This paper uses the ant colony optimization algorithm and the rough set to carry on the research,and mainly completes the following work:1.This paper introduces the concepts and theories of rough set theory,and illustrates several methods of roughness attribute reduction.It provides theoretical basis of combining with ant colony algorithm and rough set attribute reduction method using ant colony optimization algorithm(ACOAR algorithm).Then,algorithm principle and algorithm model of the basic ant colony algorithm are studied in the ant colony algorithm,and the advantages and disadvantages of the basic ant colony algorithm are analyzed.On this basis,the ant colony optimization algorithm and a hierarchical ant colony optimization algorithm are introduced.2.By using the difference matrix method and the ant colony algorithm,a new reduction method is constructed.As far as possible simplify the redundant attributes to achieve the desired goal.The algorithm compared with the original reduction method has been improved.In this paper,an improved ant colony algorithm(ACOAR algorithm)is proposed by using the ant colony algorithm.The algorithm starts from the update of pheromone and limits the upper and lower limits of pheromone values.And then the algorithm based on the addressing method to improve the candidate solution of the construction algorithm and other aspects to develope the new ant colony combination method.After the test verification and comparison,it has its unique advantages.