Benchmark datasets and real-time autoimmune disease dataset analysis using machine learning algorithms with implementation, analysis and results

Abstract

A machine learning model intends to produce a secure model with low bias and variance. Finding the optimal machine learning model for a dataset is a challenging task. A suitable machine learning model is yet to be specified for the Arthritis Profile Data dataset. Autoimmune disease is widely spread all over the world. Some autoimmune arthritis diseases are Rheumatoid Arthritis, Psoriatic Arthritis, Juvenile Arthritis, etc. These diseases come under both categories autoimmune and inflammatory diseases. The proposed work is designed to suggest the best machine learning model with the highest observed accuracy for the Arthritis Profile Data. Many authors do not compare newly created datasets with previously used datasets. This can lead to inaccurate results due to the lack of reliable comparison. Additionally, it can prevent researchers from detecting potential bias in the data. Comparing datasets can help to identify and address any potential issues and improve the accuracy of the results. It is important to review existing datasets before beginning a new project to ensure the accuracy of the results. This article is the first study on the topic that analysis the accuracy behavior of each machine learning model concerning the Arthritis Profile Data and various benchmark disease datasets with different hold-out and k-fold cross-validation methods. The study concludes with a glimpse of whether dataset and feature size affect model prediction accuracy and proffers a machine learning model for the Arthritis Profile Data. The proposed research explores base learning classification algorithms and ensemble methods such as Logistic Regression, K-Nearest Neighbor, Support Vector Machine, Random Forest, and Extreme Gradient Boosting from machine learning. Our empirical evidence clearly states XGBoost ensemble technique shows the highest accuracy for the Arthritis Profile Data.