Development of a Modular Case-Based Reasoning System for Data Analysis

Abstract

This paper is devoted to the issues of developing a modular case-based reasoning system, capable of performing all stages of the learning cycle and its implementation as a web application. The main modules of the system and an example of its usage for solving the task of data classification is presented. k-NN algorithm is used for classification. The system also used to optimize case library by reducing the number of cases in case library using various classification (based on the k-NN algorithm) and cluster algorithms (based on the k-means algorithm).