Automatic Detection of Hidden Regularities Based on the Study of Class Properties
В. Г. Родченко · Pattern Recognition and Image Analysis · 2020
Abstract— The paper presents a method for automatically detecting in data practically useful and interpreted regularities for decision making. The method uses the compactness hypothesis and analyzes estimates for the mutual placement of class patterns in decision spaces. Class patterns are represented as cluster structures constructed from training sample data. The results of applying the method to solve a classification problem based on a real data set are demonstrated.