What we want to do is determine the ESRB rating for a game given a list of content descriptors for a new game the ESRB hasn’t seen before. For example, the Nintendo DS adaptation of Chrono Trigger was rated Everyone 10+ and had content descriptors for animated blood, fantasy violence, suggestive themes, and use of alcohol.ĮSRB ratings also contain more context in their summary as pictured below: Games are classified into one of five buckets:Ĭlassification is done based at least partially by content descriptors in games. and training a classification model to predict which ESRB rating label will be applied to new games given the features of that game.ĮSRB, for those not familiar, stands for the Entertainment Software Rating Board and they classify games so that parents and other guardians can make informed decisions based on game contents when purchasing games for minors. While these are similar and may share similar code under the hood, they are two different technologies Our Classification Problemįor the rest of the article, we’ll be looking at a collection of video games with titles and whether they contain things like blood, violence, language, nudity, drug references, etc. We’ll discuss this tool more at the end of the article, but this article primarily focuses on the code aspects of Auto MLĪdditional Note: Auto ML is also sometimes also used to refer to Azure Machine Learning Studio’s Automated ML capabilities. Note: Auto ML can also refer to the interactive model generation tool that comes with ML.NET. This helps people newer to data science find a model that performs well without needing a much larger data science skillset. NET technologies to quickly get started with machine learning with their existing technical stacks and skills.Īuto ML is a subset of ML.NET that abstracts away the process of choosing a machine learning algorithm, tuning hyperparameters for those algorithms, and comparing algorithms to each other to identify the best performance. NET applications or for developers familiar with. In short, ML.NET is a great way for teams with existing. NET, ML.NET is cross-platform, but it is also extremely performant and can be used for a wide variety of machine learning tasks. ML.NET is Microsoft’s open-source library for machine learning in.