machine learning : Handwritten digits recognition - MNIST

I need to implement 3 algorithms out of the list below, on [login to view URL] ([login to view URL], 70,000 samples (total) • 10 digits (0,1,…,9) • 28x28 pixels (784-D) ). The 3 algorithms I have chosen are: 1) SVM with various kernels and hyperparameters 2) k-means 3) logistic regression For each algorithm I need to mention the hyper parameters (for instance, what is the polynomial degree in SVM and why did we choose this one. I should use cross validation and grid search) . I need to change the dimension for our dataset, hence, we need to add appropriate graphs and explain why we chose this dimension. At the end , I need to make a table that compares the 3 algorithms according to some criterions and give a score for each. It should look similar to this work ([login to view URL]) but it should be written on jupyter (.ipynb file). Remarks: -Please read the instructions above carefully!!! -The task must be fully and clearly documented!!! - See attached file (about the project) below. -Required skills: 1. Machine Learning 2. Jupyter