https://github.com/samyc2002/mnist-image-classification-task
A ML algorithm to predict the drawn number
https://github.com/samyc2002/mnist-image-classification-task
Last synced: about 2 months ago
JSON representation
A ML algorithm to predict the drawn number
- Host: GitHub
- URL: https://github.com/samyc2002/mnist-image-classification-task
- Owner: Samyc2002
- License: gpl-3.0
- Created: 2021-03-12T12:46:03.000Z (about 4 years ago)
- Default Branch: main
- Last Pushed: 2021-03-12T12:50:14.000Z (about 4 years ago)
- Last Synced: 2025-02-10T03:13:29.167Z (4 months ago)
- Language: Python
- Size: 4.26 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# MNIST Flask App
A Flask web app for handwritten digits using a logistic regression model
Used as a follow up assignment with [Chapter-1: Basics, ML Workshop, AI-Club, IIT Dharwad](https://github.com/deepakhr1999/MachineLearningWorkshop)
### Step 1: Download code
You can download this code using this link https://github.com/deepakhr1999/mnist-flask/archive/master.zip
Uncompress the zip file. The code must be in a folder called `mnist-flask-master`
### Step 2: Google drive link for mnist data
Download these files into the master folder you just downloaded
- train.csv : https://drive.google.com/file/d/1-2wqdPvCunWwcenwZwCSYFfFwk38XEXO/view?usp=sharing
- test.csv : https://drive.google.com/file/d/1-DVETsf0kvowpAx8H6EDD7ZYJO8cPPRK/view?usp=sharing
### Step 3: Install requirements
Run this command in the windows powershell
```sh
pip install scikit-image flask pandas scikit-learn
```
### Step 4: Complete the missing code
- Complete missing code in the file named LogisticRegression/train.py
- In the `mnist-flask-master` folder, Shift + right-click and open powershell. Run the following command
```sh
python LogisticRegression/train.py --savefile SavedModels/logreg.pkl
```
### Step 5: Start the flask app
- In the `mnist-flask-master` folder, Shift + right-click and open powershell. Run the following command
```sh
python app.py --model LR
```
- Open your favourite browser and go to the link http://localhost:5000
- Draw a digit and hit predict!
- The model you trained will be used to classify :)
- Expect medium performance, this is a simple model
- Way better than trying to write our own code (without ML) to identify patterns.