Digit Recognizer Python Jupyter Notebook with Convolutional Neural Network digit recognizer implemented in Keras
Data Dataset: MNIST Handwritten digits
Description: Classification of handwritten digits, 10 classes (0-9).
Training: 37.8k (0.9) images
Validation: 4.2k (0.1) images
Testing: 28k images
Model
conv2d_22 (Conv2D) (None, 28, 28, 32) 832
conv2d_23 (Conv2D) (None, 28, 28, 32) 25632
max_pooling2d_11 (MaxPooling (None, 14, 14, 32) 0
dropout_7 (Dropout) (None, 14, 14, 32) 0
conv2d_24 (Conv2D) (None, 14, 14, 64) 18496
conv2d_25 (Conv2D) (None, 14, 14, 64) 36928
max_pooling2d_12 (MaxPooling (None, 7, 7, 64) 0
dropout_8 (Dropout) (None, 7, 7, 64) 0
flatten_4 (Flatten) (None, 3136) 0
dense_8 (Dense) (None, 8192) 25698304
dropout_9 (Dropout) (None, 8192) 0
dense_9 (Dense) (None, 2048) 16779264
dropout_10 (Dropout) (None, 2048) 0
Total params: 42,579,946 Trainable params: 42,579,946 Non-trainable params: 0