With self-learning models, prediction can be explained

Self-learning computer models can be of value for speech recognition, fraud detection, and assessing medical risks. However, the benefits scandal, for example, shows that the utmost caution must be exercised: and that is why the law stipulates that there must always be an explanation as to how a model reaches a certain conclusion. To help data experts with this, Ph.D. researcher Dennis Collaris has developed interactive visualization tools that offer insight into the "thought processes" of artificially intelligent models.

computer sciences explained learning prediction self-learning