Machine learning and applications to real-world data analysis

The world around us is brimming with data, and even this text you are reading right now constitutes a part of it. By effectively combining data with human knowledge and a bit of ingenuity, there is the potential to solve real-world problems and discover new knowledge. My research interests include the development of machine learning and pattern recognition algorithms, with a focus on their practical application in data analysis to tackle real-world problems. To accomplish this, it is necessary not to simply apply existing machine learning algorithms to the obtained data but to generalize the problem that the target task involves and devise novel algorithms while incorporating domain knowledge. Specifically, my work involves the integration of domain knowledge into neural networks, devising methods to estimate the uncertainty and reliability of predictions by machine learning, and designing learning algorithms utilizing limited data. The applications range widely, including biological signals such as EMGs and EEGs, medical images such as MRIs, documents, and generic images.