Abstract: Techniques are disclosed relating to assessing technology activity using image-based machine learning algorithms. A computer system may access a data set that includes a plurality of parameters (e.g., technologies) for an item (e.g., a web-based interface). The plurality of parameters may correspond to a plurality of time intervals. The computer system may generate a two-dimensional graphical representation of the data set. A first dimension of the graphical representation may be indicative of values of the plurality of parameters at different time intervals and a second dimension of the graphical representation may be indicative of a time period that includes the plurality of time intervals. At least one characteristic of the data set may be determined by inputting the graphical representation of the data set to a trained machine learning module. The trained machine learning module may implement an image-based learning algorithm.

Abstract: Techniques are described relating to analyzing user transactions based on time of day of occurrence, and using time of a day as a factor in determining whether a new transaction should be allowed or disallowed. People may have particular tendencies to engage in transaction at certain times of a day. When a new transaction occurs that does not fit a previous pattern, this can indicate someone else has gained access to the account. Past times of transactions can be transformed to a two-dimensional representation that avoids discontinuity. A smoothed probability distribution can indicate a likelihood of whether a new transaction fits previous patterns. If a new transaction is unlikely due to time of day, the new transaction might be denied/prevented from completing. Denial of the transaction may also be based on additional factors besides the time of day.

Abstract: Techniques are described relating to identifying a country (or other item) associated with an individual based on the individual's name. These techniques rely on machine learning and artificial intelligence adaptions, according to various embodiments, and allow for better identification of country than some alternative techniques. Specifically, unsupervised machine learning techniques (e.g. using a word2vec based algorithm) allow for the handling of noisy data, which can be a significant difficulty in attempting to associate a person's name to a particular country, where it may be quite difficult or even impossible to train a supervised machine learning model that can effectively make decisions on how to associate an unknown person to a particular country. Accordingly, this disclosure includes techniques related to unsupervised machine learning that are particularly helpful for solving this problem, including using a training data set that is prepared by adding country codes (or another identifier) to names.