The weekly sales transaction dataset shows weekly sales of over 800 items across a 
year. Your task is to predict the final week's sales from the previous values for 
each item in the dataset. Report your mean squared error which is defined as the 
mean squared error of your predictions 1/n(sum_i (y'i - yi)**2). The best mean 
squared that we achieve in this dataset is about 17.5 with ridge regression applied 
to an LSTM encoding of the data.