In this section, you're going to apply your knowledge of multiclass classification to the MNIST Dataset. This is a famous dataset that represents a collection of images depicting handwritten digits.

You'll build a new training pipeline to process the data, use a multiclass classification algorithm to train a machine learning model, and then evaluate the predictions generated by the fully trained model.

Finally, we'll explore the Micro and Macro Accuracy metrics and I'll share some tips with you on how you can improve the predictive quality of your model.