Predictive Tool for Dermatology Disease Diagnosis using Machine Learning Techniques
M. Sudha1, B. Poorva2
1M. Sudha, Associate Professor, School of Information Technology & Engineering, VIT Vellore, India.
2B. Poorva, PG Scholar, School of Information Technology & Engineering, VIT Vellore, India.
Manuscript received on 28 June 2019 | Revised Manuscript received on 05 July 2019 | Manuscript published on 30 July 2019 | PP: 355-360 | Volume-8 Issue-9, July 2019 | Retrieval Number: G5376058719/19©BEIESP | DOI: 10.35940/ijitee.G5376.078919
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Prediction of skin diseases is more complex as many diseases have the same symptoms at the early stage but may vary at the later stages while the disease becomes incurable. So we can use data mining algorithms to classify the diseases based on the input symptoms. In this paper, the best algorithm suitable for classification of data into six dermatological diseases is determined by comparison with few other algorithms. Naive Bayes tends to show higher accuracy of 99.31% , Random forest exhibits 97.80% and SVM reveals 94.35% when test size is 40% in jupyter notebook. Linear regression and K Nearest Neighbors when trained with 80% of the data displays 82.14% and 94.44% accuracy respectively. Naive Bayes can be used for the prediction of several other diseases and is best for classification of data and thus helps doctors predict the disease more accurately and with comparatively lesser time.
Index Terms: K Nearest Neighbors, Linear Regression, Naive Bayes, Predictive Model, Random forest, Support Vector Machine,
Scope of the Article: Machine Learning