Machine learning harnesses the computational power of modern computers to apply complex statistical models to large amounts of data. As experimental research on nanopore sensing progresses, more and more data from nanopore sensing experiments on proteins in their native conformation become available. The goal of my project is to identify the right signal characteristics and the right statistical models that would allow us to extract a maximum of information from nanopore sensing data. Statistical models such as random forest could then detect correlations between signal characteristics and protein type, which would enable us to classify unknown proteins based on nanopore sensing data.