This project was carried out as part of a competition on Kaggle.com. I have implemented an image classification algorithm to predict human handwriting. For the modeling, I used a convolutional neural network (CNN) with five hidden layers and a customized optimizer in order to avoid local minima problems and push the algorithm to converge more quickly. The architecture of the algorithm briefly explained just above allowed to get a 99.2% of accuracy !!
-
Notifications
You must be signed in to change notification settings - Fork 0
khaledyac/Digit_Recognizer
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
An image classification algorithm to predict human handwriting
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published