DiscimusRW: An E-Learning Web Application for Classifying Random Walks with Machine Learning

Abstract

Random walks are key for many fundamental processes, including the diffusion of substances in solvents, epidemics’ spread, and financial markets’ development. Machine learning applications have not only revolutionized research in various areas, such as image processing and protein structure prediction, but are also abundant in daily life. For example, machine learning enables us to develop self-driving cars and helps us to improve online shopping suggestions. Both random walks and machine learning are often taught theoretically in schools and universities because of the lack of means to provide experiential learning opportunities. However, the absence of experiential learning may cause low educational attainment rates of students in courses that address these topics. Here, we discuss a web-based online system based on Django web technology to support teaching random walks and machine learning. The application uses machine learning models to classify 2D random walks that a user can provide. In a teaching setting, the application can—for example—be applied to determine the sizes of spherical particles that diffuse based on their trajectories.