Emir U 5/12/2023

Semi-supervised clustering with logic programming

Read Original

This technical article argues that logic programming, specifically Prolog, is well-suited for general clustering tasks, especially when domain knowledge is expressed as rules rather than numerical vectors. It critiques traditional algorithms like k-means for their limitations with non-commensurate data and proposes artisanal Prolog clustering algorithms as a more intuitive, rule-based alternative for scenarios like customer segmentation.

Semi-supervised clustering with logic programming

Comments

No comments yet

Be the first to share your thoughts!

Browser Extension

Get instant access to AllDevBlogs from your browser

Top of the Week