A Hybrid Application Model to Improve the Accuracy by Using Naive Bayes Algorithm

In this paper, we mainly focus on improving the accuracy in the joining rate of values. The dataset comprises of one lakh records with different features as input and further performed data cleaning using data profile report for handling the missing values and unnecessary data. After performing pre-processing and cleaning the data, an effective model is built using five different algorithms based on which we propose the implementation-based predictive model with more than 75% accuracy. The selection process is performed by achieving maximum conversion rate over selective targeted clients by denoting the maximum probability of each lead that enrolees in the university through consultancy with possible discounts or contacts with high potential leads. The training set is used to train the classifier, and test data is used to generate a confusion matrix by calculating the accuracy. Several techniques can be used to solve this problem. The proposed Naive Bayes algorithm has achieved an accuracy of approximately 93.5% on the testing data, which is greater than decision tree accuracy.