Automated Detection of Biases within the Healthcare System Using Clustering and Logistic Regression

Abstract

Big data play a vital role in decision-making, especially in the healthcare system such as in cardiovascular disease. However, the findings of algorithms used in decision-making show some disparity as compared to the existing findings of physicians. This is due to the biases in the big data set used for the healthcare system. This will lead to misdiagnosing certain protected groups or attributes like gender. Therefore, it is the major problem to detect biases in a large dataset. In this paper, we have proposed a model and implemented it to detect biases in the large data set of cardiovascular disease. This model uses statistical performance metrics to measure the biases in the dataset. The result shows that if we apply the clustering mechanism with logistic regression along with statistical performance metrics it gives a better result to detect biases in the dataset.