Alphabet Recognition ⚓

This code helps you classify different alphabets using softmax regression (lower case).

Code Requirements 🦄

You can install Conda for python which resolves all the dependencies for machine learning.

Description ⚙️

Softmax Regression (synonyms: Multinomial Logistic, Maximum Entropy Classifier, or just Multi-class Logistic Regression) is a generalization of logistic regression that we can use for multi-class classification (under the assumption that the classes are mutually exclusive). In contrast, we use the (standard) Logistic Regression model in binary classification tasks.

For more information, see

Python Implementation 👨‍🔬

1. Dataset- Extended MNIST dataset (letters)
2. Images of size 28 X 28
3. Classify alphabets from a to z
4. Logistic Regression, Shallow Network and Deep Network Support added.

Results 📊

Execution 🐉

To run the code, type python Alpha-Rec.py

python Alpha-Rec.py

📌 Cite Us

To cite this guide, use the below format:

@article{Alphabet-Recognition-EMNIST,
author = {Bahadur, Akshay},
journal = {https://github.com/akshaybahadur21/Alphabet-Recognition-EMNIST},
month = {01},
title = {{Alphabet-Recognition-EMNIST}},
year = {2018}
}