Skip to content

Latest commit

 

History

History
4 lines (3 loc) · 357 Bytes

README.md

File metadata and controls

4 lines (3 loc) · 357 Bytes

MNIST-CNN-Example

The MNIST dataset is a set of 28x28 drawn images of numbers between 0-9. A common exercise is to use convolutional neural networks (CNNs) to create a model capable of identifying the numbers pictured.

Using a model trained for 20 epochs on 60000 training images, I created a model that can identify numbers drawn with >99.7% accuracy.