Clustering Algorithms for Economic and Psychological Analysis of Human Behavior

Анотація

This article proposes an approach to analyze the behavior of groups of people and shows how to predict person location for the next month. The clustering algorithms were used in this research. Also was inspected the problem of finding associative rules. We used scalable Apriori algorithms to find the best rules. For analysis, we used the standard mlxtend library to aggre- gate data by cluster, user login, and time. In this article we explored apriori and k-means clustering algorithms to get user behavior analysis template. In the process, we looked at the problem of finding associative rules that were able to find and describe patterns in large datasets. We used scalable Apriori algo- rithms to find the best rules. For analysis, we used the standard mlxtend library to aggregate data by cluster, user login, and time. While working, we were faced with the problem of inaccuracy and inconsistency of data with real condi- tions, and were forced to reduce the minimum support for associative rules.

Класифікація

Ідентифікатори

Схожі роботи