Predicting Life Expectancy of Acute Myeloid Leukemia ( AML ) Patients Based on Gene Expression of Cancer Cells

Doctors must often decide whether to offer a cancer patient a risky treatment or let the patient continue to live with cancer. Trialing a variety of dimensionreduction techniques, we determine one that best clusters cancer patients based on survival data. Then we predict the life expectancy of a cancer patient using a KaplanMeier estimator on reduced-dimension data. PCA was the most effective dimension-reducing algorithm based on our data visualization metric, and was used to show that reduced data can be used to predict the survival rate of cancer patients.