Abstract:
In the world millions of people suffer from hemorrhoids. It is a medical condition that
frequently requires prompt intervention for effective management. In this paper, we
suggest a machine learning-based method for estimating an individual's risk of developing
hemorrhoids. We are using a dataset that includes lifestyle factors, medical history, and
demographic data. We use five algorithms: K-Nearest Neighbors (KNN), Support Vector
Machine (SVM), Random Forest, XGBoost and Naive Bayes to create prediction models.
Cross-validation and feature selection strategies are applied to improve the validity and
performance of the model in this paper through dataset. Our research shows encouraging
outcomes in precisely identifying people who are susceptible to hemorrhoids, which will
enable early interposition and individualized treatment plans. This work offers a proactive
approach to hemorrhoid management and its prevention system, furthering the field of
predictive analysis in healthcare. Random Forest’s accuracy 95.19%, KNN’s accuracy is
95.21%, SVM’s accuracy is 95.93%, XGBoosting’s accuracy is 95.32% and Naive Bayes’s
accuracy is 90.37. SVM’s training accuracy is the highest and it is 96.67% So, SVM is
suitable and best algorithm for predictive-probability analysis and preventive strategies for
hemorrhoids. We have 1350 primary data. The dataset consists of 670 of patient records
and 680 of normal records. Additionally, a brief summary of preventive strategies intended
to lessen the frequency and severity of hemorrhoids is included in our research paper. This
paper seeks to equip people with the knowledge and skills needed to reduce the burden of
hemorrhoids and improve their overall quality of life by spreading awareness about
preventive measures.
