`semi_conductor.csv` contains the semi_conductor quality-control data from a semi-conductor manufacturing process. There are hundreds of diagnostic sensors along the production line, measuring various inputs and outputs in the process. `codebook_semi_conductor.txt` contains description of the dataset.

(1) Do in-sample-fitting by using random forests.

(2) Do in-sample-fitting by using factor models.

(3) Compare the R-squared from parts (1) and (2).

(4) Which model does best in predicting unseen data? (Hint: evaluate the out-of-sample performance of each model through k-fold cross valuation)