Web-based Interface for Breast Cancer Prediction for classification of benign and malignant using Machine Learning.

Abstract

This paper emphasizes directions for future research and various gaps in
research for a better understanding, accuracy, reliability, and clinical
implementation of breast cancer diagnostic systems. This web-based UI will
allow doctors to upload patient data to get instant predictions based on selected
algorithms. A final AI pipeline, from preprocessing of the data to the
deployment of the model, will be shown by this system, so that it will target to
help to find early diagnosis of breast cancer.
Worldwide, among cancer-related deaths, it has become one of the prime
causes in females, contriving early and precise diagnosis is a serious challenge
in healthcare. Orthodox diagnostic techniques depend deliberately on expert
interpretation and are arduous. To deal with such restraints, data–centric task
and machine learning–based approaches are increasingly adopted to support
clinical decision-making. Due to its usefulness and clinical applicability, the
Wisconsin Breast Cancer Dataset has become a broadly used benchmark for
measuring computational models for tumour classification. This review
critically analyses the present research on the diagnosis of breast cancer,
leveraging this dataset, employing statistical methods, techniques for classical
machine learning, accumulated models, and a deep learning approach. The
survey contrast report performance speech analyses, systematic practice,
conversation challenges, constraint and future research methods in checkup
artificial intelligence.