Analysis of Different Algorithms to Predict Chronic Kidney Disease

Chronic kidney disease (CKD), also referred to as chronic kidney failure, is a medical condition defined by a gradual loss of kidney function over a period of time. It is characterized by conditions that inhibit or worsen the ability of the kidneys to filter wastes from human blood. If the disease worsens, wastes could build up to alarming levels in our blood and lead to organ failure, after which dialysis or kidney transplant would be needed to maintain life. It could further cause complications such as high blood pressure, low blood count (also known as anaemia), reduced bone strength, poor health and nutrition, and neurological damage. Since these symptoms usually slowly develop over time, early diagnosis can help chronic kidney disease from getting worse. This paper aims to use a set of medical attributes and analyse different machine learning algorithms to develop a model that would help predict whether or not a patient has chronic kidney disease. It also aims to perform a comparative analysis of multiple classification algorithms along with ensemble stacking method by testing their performance.