Text-Classification

Take five different samples of Gutenberg digital books (or your choice of text corpus), which are of five different authors, that you suspect are of the same genres and are semantically the same. For example, choose two of the books 1- The Brothers Karamazov and 2- Thus Spoke Zarathustra. ...5-

Separate and set aside unbiased random partitions for training, validation and testing.

The overall objective is to produce classification predictions and compare them; analyze the pros and cons of algorithms and generate and communicate the insights.

Gauge the bias and variability of the models to decide the champion model.

Then play with the features and other factors that provide you with leverages to make it harder for the model to predict and bring the accuracy down for about 20% and then check the bias and variability.