/Predicting-global-game-sales

Predicting global game sales with R

Primary LanguageHTML

Predicting global game sales

The project has two goals: a) to demonstrate how to align tidymodel codes for regression analysis, and b) to demonstrate how good the model is at predicting outcomes. The ultimate aim of the study is to investigate whether or not an accurate Machine Learning model can be built to forecast video game sales in units based on the features given in this dataset. This hypothesis is investigated with numerous supervised ML models.

The data comes from the Kaggle. Motivated by Gregory Smith's web scrape of VGChartz Video Games Sales, this data set simply extends the number of variables with another web scrape from Metacritic.

Basically, this analysis includes.

  1. Data manipulation
  2. EDA analysis
  3. Tidy style model and workflow creation.
  4. Tunning.
  5. Identifying the predictor importance.
  6. Model validation.

To check the results, go to MD files or click here.