Survey-based Machine learning approaches to diagnosis of hair fall disorder in Bangladeshi Community
Farhana Khatun, Moshfiqur Rahman Ajmain, Sharun Akter Khushbu, Nushrat Jahan Ria, Sheak Rashed Haider Noori
Abstract
Hair symbolizes the beauty of women and men. All of us are jealous of our hair. We lose hair at a young age due to some mistakes or irregularities. Lots of men and women all over the world are suffering from hair falling and the number of females is suffering growing per year. Genetically, dandruff, allergy and stress are the major problems for falling hair. We are doing this research survey for helping people. This study is representing two things. First of all, we are findings how many reasons are involved in hair fall. Another thing is we train our dataset with machine learning algorithms to find out the accuracy. Machine learning technologies have rapidly evolved to analyze survey datasets. SVM, Logistic Regression, Naive Bayes, Decision Tree, Random Forest, K-nearest Neighbor and XGBoost algorithms for performance comparison. The experimental results indicated that XGBoost had the best performance, with an accuracy of 92.62%.