This is coursework that I tackled as part of my 3rd year at university. It involved predicting a person's oral temperature based on the readings of a thermometer. The device had two modes to capture the person's temperature: the fast mode and the monitor mode. Our job was to develop the best model to identify whether the person is ill or not. The number of models we had to develop was not specified nor was the type of models we had to build (both regression models and classification models were acceptable - while "ill" or "not ill" can be represented as boolean values, you could also predict the person's temperature using a regression model and then use a threshold to determine whether they are ill or not).
The Jupyter Notebook walks you through the steps I took to determine the best models for the tasks at hand.