DEVELOPMENT OF A LEARNING ADAPTABILITY SYSTEM FOR MTU STUDENTS USING A DEEP LEARNING ALGORITHM

Abstract

The aim of this study is to adopt the use of deep learning algorithm for the development of a learning adaptability system which can be used for classifying students based on relevant information. The specific objectives are to identify existing studies, construct instrument of data collection, collect relevant data analyse the data collected and develop the model, implement the system based on the results and test the system. The study identified the various user and system requirements, specified the system design, and implemented the system. A review of the literature was being done to identify and understand existing works, a structured questionnaire was constructed according to the Felder-Silver Model for collecting data from students of MTU, relevant data was collected from 600 students using a simple rando sampling technique, the collected data was analysed using a deep neural network architecture and the system was implemented using python. The results of the system showed the implementation of the system‟s database with the use of data mining algorithm for the extraction of features from external environment and classification in order decipher whether a student learning style. Other systems have been built already, but some of these existing systems are expert systems, and are often complex and hard to relate with. Furthermore, they are not so accurate in their prediction and hence are not so reliable. The system design was specified using UML diagrams, such are use case, sequence, and class diagram