Thesis

Presenting the Fuzzy Cluster-Based Decision Making Model Based on Machine Learning

(Maintenance Case Study)

Abstract

In my thesis, a new multi-criteria decision-making framework is presented. This framework is formed by combining machine learning, fuzzy numbers and multi-criteria decision-making and has reduced the number of alternatives, increased decision-making speed, and reduced decision-making uncertainty.

In the proposed method, the data is first pre-processed and then assigned to several clusters using clustering techniques, where each cluster will be a new decision alternative. Clusters' features were expressed with fuzzy numbers to be influenced by all cluster members and reduce the uncertainty.

Finally, using multi-criteria decision-making, the clusters are ranked, and the priority of each cluster is determined. This method was particularly used to prioritize IoT applications in predictive maintenance. As a result, 201 decision options were divided into 7 clusters, and the decision operation was carried out. The specified clusters were presented in the order of priority of implementing the Internet of Things.

Google Sites

Report abuse