Outcome Based Predictive Analysis of Automatic Question Paper Generation Using Hybrid MLP-CNN Approach

Improving the overall caliber of the student body is the overarching goal of any educational area. In order to save teachers time and effort, we implement cutting-edge technology in the form of a system that can automatically produce questions for exams. This was due to the fact that each writer's knowledge, experience, and individual tastes contributed uniquely to the final product. This approach eliminates the need for human work by automatically processing unique paper collections. In a second, the computer system picks up all the paperwork and tedious responsibilities. Database of questions categorized by category, course, chapter, and number of points awarded. Exam papers can only be reliably produced by the automated paper generator. This system automates the entire paper-making process. For instance, picking a question bank to work from could be the first step in writing a paper. This approach works for organizations of all sizes. The proposed approach employs the procedures of preprocessing, feature selection, and model training. Preprocessing methods often make use of missing values. With its roots in subsampling and selection algorithms, stability-based selection is an innovative approach to deciding on characteristics. Once features have been decided upon, Hybrid CNN-MLP is utilized to train the models. In contrast to the two most used substitutes, the proposed technique outperforms both CNN and MLP.