Classification of learning styles in a personalised multi-agent system using machine learning frameworks for stem education

The learning style and distinct needs of each student cannot be fulfilled through the traditional teaching methods where personalized adaptive learning helps user to understand, retain and monitor its framework based on its subjective needs as discussed in Nouman et al. [1].The main aim of intelligent systems in a multiagent framework is to classify and learn various methods and styles accurately that is discussed in Ni, et al. [2].It uses Visual, Auditory and Kinesthetic in a STEM based approach which personally analyze the contents enhancing the outcomes through the feedback from the learners as discussed in Ayyoub and Al-Kadi [3].we classify each learners style through student inputs. In proposed method, we use variety of multiagent systems such as Random Forest (RF), Decision Tree (DT), Support Vector Machine (SVM), and Multinomial Naive Bayes consisting of four agents: Teacher, Concept Mapping, Content Analysis, and Sentiment Analysis. Here, Multinomial Naive Bayes Framework achieves best results with accuracy of 98% where insighted analysis on sentimental and engagement level of disabled is discussed in Chang and Lin [4]. In this study, the personalized individual learning styles are promoted and engaged in STEM education where the expansion of the various learning styles pays the way for a student-centered approach in real time involved in Gonçalves [5].