A) Basic Model : Use data features after preprocessing Evaluating all Models
B) Rank Based Model: Considers popularity of one news item affecting share value of other news items. Evaluating all Models
C) Dimension Reductionality + Rank Model - Dimensionality reduction used along with the rank model for prediction. Notebook
D) Under Sampling + Dimension Reductionality + Rank Model - added under sampling for the news item with 0 shares as there were very high number of such items. Notebook
E) Time Series Based - Used Recurrentmd Neural Network LSTM cells. Notebook