Img_compression_K-Means

Using K-Means clustering I reduced the size of an image. Where the input image may has more than 16M possible colors, the output image only consists of 16 colors.
Here I used some interactive plot so that the user manually choose the no of 'k' in the algorithm using a slidebar. When the mouse will be released, it automatically train the model and plot 2 image side by side. The plot_utils consists of helper function that I used to ploting. Some important concepts and intuition is added in the notebook, (little theory). Although it is not a good way to compress image. I just showed how to do that using ML.
The source code in python is added. I also include some image on which I performed the algorithm(only 4).