Hybrid Comparative Predictive Modeling

In this research, a hybrid predictive model was proposed for the language assignment decision-making process where predictive modeling is primarily based on personality type and learning style preference attributes. Predictive models can be built through the use of machine learning using different classifiers/algorithms to predict results as well as provide recommendations to management for placement in appropriate language programs of study and students for the adoption of appropriate study strategies and habits. A predictive model through machine learning was used in conjunction with probabilistic classification and clustering of specific segments within the data in order to increase the rate of success for an improved decision-making process. Variance in the actual and predicted results with respect to the difference in success rates can assist the decision makers in student placement. An aggregate of all the processes with the help of Cobb-Douglas utility function leads to a Hybrid Predictive Model, which combined two different phases for better placement, an increased rate of success, and an overall improved decision-making process. The introduction of Cobb-Douglas utility function can further streamline the process to check any external factors that may have influenced the predicted results.