Social spider optimization algorithm for effective data classification: An application of stock price prediction

Abstract

Presently, data classification has become a hot research topic, intends to categorize the data into a predefined number of classes. The data classification problem has been considered as a NP hard problem and various optimization algorithms are introduced to solve it. Social spider optimization (SSO) algorithm is originally developed to resolve continuous and optimization problems. In line with, it has been altered to manage different kinds of optimization process and also to be employed for data investigation. And, some other studies are also investigated the use of SSO algorithm in different domains. In this paper, we introduce new SSO algorithm particularly applicable for data classification process. In order to validate the performance of the SSO algorithm, a real time problem of stock price prediction (SPP) is employed. For experimentation, the results are validated by testing the SSO algorithm against four datasets such as Dow Jones Index (DJI) dataset, three own datasets gathered from Yahoo finance on the basis of daily, weekly and yearly. The empirical result states that the proposed algorithms perform well and it is noted that the classification performance of the SSO algorithm better than the compared methods.