Prognostication of Methicillin-resistant Staphylococcus Aureus (MRSA) patient survival

Prognostic methods are potentially beneficial for public health management. The blending of data-driven methods with the domain knowledge is essential to efficiently advance feature selection, anomaly detection, prognostics forecasting, data matching and clustering. This paper attempts to demonstrate how prognostic methods enable accurate Methicillin-resistant Staphylococcus Aureus (MRSA) patient life prediction. The methodology is applied to MRSA patient survival analysis. Significant linear relationship is found between log (hazard) and age (p<;0.001). By adjusting the time-depending effect of age, we construct more accurate Cox's proportional hazard models. It is believed that understanding age effect on MRSA patient survival is able to receive more robust result using prognostic approaches. To further enhance model prediction power, it is suggested to explore statistical data transformation and adjustment under various attributes.