A deep learning model to predict handwritten Kannada digits.
Dataset can be found at Kaggle. https://www.kaggle.com/c/Kannada-MNIST The data is in .csv format.
The data from the .csv files is extracted and converted into images of size (28x28).
The DL model with Convolution & Dense layers, had achived 98% accuracy on the training data & 72% accuracy on validation data.
