Prediction of Accuracy of Datasets by Using Machine Learning Algorithms on Food Habits among College Going Students in Eastern India

The eating habits of fast food may lead to various disorders such as obesity, diabetes, etc. among adults.The present study was evaluated to predict accuracy performance of dataset for food habits viz.The datasets were used as Food_habit, Number_of_meals/day, Omit_any_meal, Take_any_special_food_or_not, Type_of_food_normally_taken, Type_of_meal_preferred, Frequency_of_eating_outside, Habit_of_skipping_breakfast and class effects viz. Normal and abnormal category among college going students (35 nos.) of eastern India. In this study, the prediction accuracy was obtained through 7 machine learning (ML) algorithms especially BayesNet (BN), NaiveBayes (NB), logistic regression (LR), Stochastic Gradient Descent (SGD), Sequential minimal optimization of Support Vector Machine (SMO), K -nearest neighbour (IBK), and Lazy. KStar (K*), by using ML tool (WEKA, version 3.8.5) as per cross -validation (CV) test for above -mentioned classes. As per the statistical summary results, the prediction accuracy through precision recall curve (PRC) were obtained as highest values (98%) for these algorithms. Future study should be performed with other big dataset with ML algorithms especially Tree algorithms.