Your task is to write a program (in any language you like, preferably Python) to eliminate the noise added to this dataset. Once you have the cleaned data, upload to the following Kaggle contest to obtain the MSE between your answer and the uncorrupted training images (half of the total dataset). You may submit to Kaggle a limited number of times. This is an open-ended problem with no “right” answer. Your solution will be scored entirely by the mean square error between the file that you submit and the uncorrupted test images (the remaining half of the dataset).