Risk of Dental Disease Prediction Using Machine Learning

Abstract

Now a day’s dental disease is the major health problem in Bangladesh. So dental care is important to most people in our country. But the cost of dental care services is increasing day by day. We will predict the risk of dental disease with machine learning. We identify the most common disease among people, consult with a dentist about those diseases, reading-related journals, and online articles, we find out the habitats that cause dental disease. Then we collect data based on those factors, such as age, brush before sleep, brush after eating morning, eating chocolates, soft drinks, betel leaf/nut, etc. We collect data from both those who have already a disease and those who don’t. We have two outcomes. One is ‘Yes’ meaning they have dental disease and another is ‘No’ means they don’t have dental disease. We apply machine-learning algorithms to our processed dataset. Recently machine learning, artificial intelligence, and deep learning used in various predictions and detection systems. We use k-nearest neighbor (KNN), logistic regression, support vector machine (SVM), naïve Bayes, random forest, adaptive boosting (ADA boosting), decision tree, multilayer perceptron (MLP-ANN), Linear Discriminant Analysis (LDA), and gradient boosting classifier. In our work, we use those factors answer as input and after processing and applying the algorithm, we find out addicted or not addicted as our output with the accuracy of 95.89% on the logistic regression algorithm.