When I was a novice in Machine Learning, I couldn’t figure out where to, and how to start the model training? Even though I knew all the key points, still it was difficult for me to step into coding and I think its quite normal for every newbie.So I decided to write about the ML workflow to solve the so-called puzzle. In this article, we will work on Keras’ built-in **MNIST **image dataset. So let’s start.

1. Data Preparation:

a) Convert the Data into Numpy Arrays:

• Before feeding the data (image or text) into a model, it should be converted into Numpy arrays first. If you are just importing the pre-loaded data from an ML library like Keras/Tensorflow 2.0 you are good to go. As the data already comes into the tuple of Numpy arrays.Then, break the whole arrayed dataset into **training **(training samples+ training labels) dataset and testing (testing samples+testing labels) dataset.

#learning-to-code #machine-learning #technology #learning #artificial-intelligence