Mid-Semester Review

[austonli]

They are creating a model that inputs various movie criteria to predict IMDB scores, which is a score aggregate site indexing TV shows and other cinematography.

Pros:

1. They selected appropriate visualizations for their data, for example the rating vs budget scatterplot demonstrates density as well.
2. The report is well constructed and organized
3. They are aware of potential fit issues, and have generated several methods to account for this.

Improvements:

1. There is only a linear model, and this model needs a lot of development.
2. A potential interesting approach is using text mining to create a quality of movie title index. I do believe that the movie title does play a significant role in the kind of score it warrants.
3. There needs to be more forms of validation other than just having a low R^2 value.