Unraveling the Multifaceted Nature of Breast Cancer: Harnessing the Power of Machine Learning and Statistical Models for In-depth Investigation, Prognosis, and Treatment Optimization

Abstract

The most prevalent form of cancer in all of human history. Women typically have a shorter lifespan than men. The death toll from breast cancer is higher than that of any other type of cancer. After reaching puberty, women all around the world are at risk for developing breast cancer, however, the disease is more prevalent in elderly individuals. The death rate from breast cancer did not significantly change from the 1930s through the 1970s. In the 1980s, countries that were healthcare conscious and had early detection systems, as well as a variety of treatment modalities to eliminate invasive diseases, began to see an increase in their overall survival rates. This article employed decision tree models and time series models to anticipate breast cancer using time series data from 1930 to 2019. The breast cancer cases that occurred in the United States of America served as the basis for this study. The most precise prediction models will be produced as a result of this data analysis. In order to evaluate how well the predictions were made, we determined how accurate they were by calculating the symmetric mean absolute percentage errors (sMAPE), measured mean absolute percentage errors (MASE), and mean absolute percentage errors (MAPE).