A fun way to practice adding two numbers together where users can write their answer on a black canvas to be verified against the correct answer.

The TensorFlow model for predicting the handwritten digits was trained in batches of 1000 with two hidden layers, 50 epochs and a learning rate of 0.001. The model’s accuracy on the test set was 97.58%.

Link to GitHub Code.

Link to demo.

Technologies used include TensorFlow, Python 3, JavaScript, Pandas, Numpy, Scikit Learn.

Game States

Empty State

At the start of the game, the user is shown an empty state where there are no plants in the background.

Midgame State

Throughout the game, different coloured plants will appear in the background when questions are answered correctly. Incorrect answers result in a gradual removal of plants from the background unless the background is already barren.

Winning State

After six consecutive wins, the background will be completely populated with plants and the user can restart the game.