Hands On: Minesweeper
If you’re up for a challenge, here’s an optional exercise for you: modify the MNIST classifier to run on the Sonar dataset.[15] The Sonar dataset (also known as the “Mines vs. Rocks” dataset) contains the patterns generated by bouncing sonar signals off two different types of objects: metal cylinders (which could potentially be mines) and rocks. See if you can train the classifier to tell apart mines from rocks.
Don’t underestimate this little project. If you’re not steeped in Python, it might take you a few hours to write the Sonar equivalent of the mnist.py library, and you’ll probably need to browse the documentation of both Python and NumPy. Only tackle this exercise if you like the idea of spending some time on it. ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access