![]() Certainty has to do with the confidence you have in the result of your model. It tells you how often the model is right (for classification tasks). ![]() ![]() Accuracy is a metric used to score models. I would like to note that what I mean by “certainty” is not the same as model “accuracy”. I do not know how to quantify the decrease in “certainty” due to this input estimate. The consequence of making an estimate as part of the input to the model is that the model “certainty” has decreased. With some market research, however, you can make estimates for those input values. Some of these input variables, for example: number of North American Players, are not known for certain before the launch of a game. Players in : North America, Europe, Japan and “Other”.The model I did create predicts metacritic rating based on the following variables: I spent a brief while researching methods to obtain the data I needed, but I decided not to pursue those methods yet. I needed more data about the development stages of the video games to make the ideal model. In other words, I had data that would not be available for new predictions I would want to make. This is because the information collected in this data set was collected after the launch of the game. As I explored my data however I realized that I couldn’t achieve this result because of “leaky” data. The model I wanted to create was one that could predict the ratings of games that have not come out yet. I used a dataset from Kaggle regarding video game sales, and ratings in order to generate my model and explore the relationship between features of a game, and critic score. I decided to try to create a model that could predict metacritic scores for unreleased video games. These scores are important to the developers, and audiences alike, albeit for different reasons. In the video game industry metacritic scores can sway audiences, pushing them to try games that they might not have tried before. Averaging many scores, ideally, increases the certainty in a score. Metacritic scores are aggregate critic scores. One of the heuristics used by consumers in their quest for a great experience is the trusted metacritic score. With all of these options comes the problem of “what to pick?”. The amount of content you can consume in any medium is mind boggling, with more content being created every day. There are endless options for entertainment.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |