User authentication acts as the first line of defense verifying the identity of a mobile user, often as a prerequisite to allow access to resources in a mobile device. Risk-based user authentication based on behavioral biometrics appears to have the potential to increase mobile authentication security without sacrificing usability. Nevertheless, in order to precisely evaluate classification and/or novelty detection algorithms for risk-based user authentication, it is of utmost importance to make use of quality datasets to train and test these algorithms. To the best of our knowledge, there is a lack of up-to-date, representative and comprehensive datasets that are publicly available to the research community for effective training and evaluation of classification and/or novelty detection algorithms suitable for risk-based user authentication. Toward this direction, in this paper, the aim is to provide details on how we generate datasets based on HuMIdb dataset for training and testing classification and novelty detection algorithms for risk-based adaptive user authentication. The HuMIdb dataset is the most recent and publicly available dataset for behavioral user authentication.