Providing dimensionality reduction in classification tasks using the proposed algorithms IPSO, PSO

Providing dimensionality reduction in classification tasks using the proposed algorithms IPSO, PSO

SHima SHafiee1 Nahid Nakhaee2

1) Master of Computer and teacher education
2) Master of Computer and Applied Science lecturer,khorasan

IOI Code : XYBD-ZKEZG
Publication : 4th International Conference On Research Science And Technology(4rstconf.com)
Abstract :
    in some critical cases, the extraction and selection of outstanding features, has solved many problems. In fact, in order to extract the most valuable subset of data in a problem, dimension reduction methods are introduced. The objectives of this study to develop a new approach based on particle swarm optimization algorithm and its improved version is to solve the problem of dimensionality reduction. For the current systems to reduce the limits, such as non-availability of data efficient, large volume data, less accurate classification of the data and unexplainable loss of their value, are facing.nearest neighbor Classification reduced. The proposed techniques were able to extract optimal efficiency and high convergence speed compared to other existing techniques more efficient to operate.

Keywords : database uci algorithms pso and ipso categories nearest neighbor method (knn) issue category reducing the size