KSI TRANSACTIONS ON KNOWLEDGE SOCIETYKNOWLEDGE SOCIETY INSTITUTE ← BACK TO CURRENT ISSUE

Research Article · IT Management

Application of combined machine learning methods to boost AI creative thinking in IT Companies

Viktor Hristov

PhD student at University of National and World Economy, Bulgaria

KSI Transactions on Knowledge Society · Vol. XVII · No. 1 · pp. 22–25 · 2026

↓ Download article PDF How to cite ↓ ← Back to issue

Abstract

Machine learning (ML) is widely used in contemporary IT as vital part of AI (Artificial Intelligence) systems. The aim of the article is to make a systematization of main learning approaches. The goal is to explore different combinations of methods and their possible impact on overall AI performance from creative point of view. It does not include new or adjusted learning methods but holds its focus on the implementation process. It is mainly conceptual proposal that provides guidelines on how to leverage creativity via different learning techniques.
Using one or another method can vary depending on the purpose aimed at achieving. Different approaches to AI learning provide different results. More novelty and uniqueness in the output results do not mean certainly better creativity. New ideas must be applicable in current or perspective company product portfolio. That means there are some boundaries that must be respective during the creative process. Management of learning techniques is critical to achieve those goals - efficient AI creative results with effective practical application, and its proper adjustment would lead to better results.

Keywords: Artificial Intelligence, Learning Techniques, Creative Thinking

Subject classification

JEL CLASSIFICATIONM15, O31
UDC CLASSIFICATIONTo be verified by the editorial office

How to cite

APA 7 — bibliographic citation based on the published issue

Hristov, V. (2026). Application of combined machine learning methods to boost AI creative thinking in IT Companies . KSI Transactions on Knowledge Society, 17(1), 22–25.

Full text and references

Download the individual article PDF to read the complete methods, results, discussion, conclusions and bibliography. HTML full text and a structured reference list may additionally be provided.

Open article PDF ↗

Article history and integrity notices

Publication status
Published. Any correction, expression of concern or retraction must be displayed here with a dated formal notice. Article history dates should be added only after verification against editorial records.
Received / accepted / online published
Display after checking the editorial records.
Correction / concern / retraction
Display dated notices prominently and link to the formal record if applicable.