https://github.com/om-kanabar/sciencefair2025
This is my project for Chicago Public School's student science fair 2025.
https://github.com/om-kanabar/sciencefair2025
chicago-public-schools matplotlib-pyplot neural-networks numpy python science-fair tensorflow
Last synced: 15 days ago
JSON representation
This is my project for Chicago Public School's student science fair 2025.
- Host: GitHub
- URL: https://github.com/om-kanabar/sciencefair2025
- Owner: om-kanabar
- License: other
- Created: 2025-10-22T22:32:52.000Z (27 days ago)
- Default Branch: main
- Last Pushed: 2025-11-03T04:18:10.000Z (16 days ago)
- Last Synced: 2025-11-03T06:10:50.034Z (16 days ago)
- Topics: chicago-public-schools, matplotlib-pyplot, neural-networks, numpy, python, science-fair, tensorflow
- Language: Python
- Homepage:
- Size: 1.11 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Science Fair 2025
### By Om Kanabar
## LICENSE NOTE
Please review the LICENSE file to understand the terms and conditions for using this project.
## TOOLS NOTE
This project was made using [Visual Studio Code](https://code.visualstudio.com/) on **MacOS** *Tahoe (26.1)* with a **2022** *M2* MacBookAir with 16 GB of RAM, times may vary with different computers.
---
## A bit about the structure of this project
```ScienceFair2025/Learning```, the learning folder contains practice scripts and experiments, showing the process of exploration and how the final project was developed from these experiments.
```ScienceFair2025/Data```, where data (other than EMINST ByClass) is stored.
```ScienceFair2025/Scripts```, where the non-neural-network essential scripts are stored.
```ScienceFair2025/Models```, where the best models are stored.
---
## Goal of Project
The goal of the project is to make a lightweight convoluted neural network in Python for recognizing handwritten characters (A-Z, 0-9). It will be trained on the EMNIST Byclass dataset.
---
## Overview
## Experimental Design (draft)
- Control group: 3×3 kernel
- Experimental group: 2×2, 4×4, 5×5 kernels
- Train 3 networks per kernel size
- Run 3 trials per network
- (Results will be logged in CSV)
2. **DV:** Model Performance Score = `accuracy / (accuracy + (0.75 • inference_time))`
---
# Instructions for Rerunning the Project
## ERROR NOTE
If any errors are found please create an issue at: https://github.com/om-kanabar/ScienceFair2025/issues/ if you are not able to do that please contact Om Kanabar at https://omkanabar.com/contact
## Before Starting
### Installing Requirements Before Starting Project
Before running the project, make sure you have Python 3 installed.
1. Open the terminal **in the project folder**:
- In Finder, right‑click the project folder and select **"New Terminal at Folder"**, *or*
- Open Terminal and type `cd`, then **drag the ScienceFair2025 folder into the Terminal window** and press Enter.
2. Create and activate a virtual environment (optional but recommended):
```python -m venv venv```
```source venv/bin/activate```
3. Install all required packages:
```pip install -r requirements.txt```
### Check for corrupted data
1. Open the terminal **in the project folder**:
- In Finder, right‑click the project folder and select **"New Terminal at Folder"**, *or*
- Open Terminal and type `cd` , then **drag the ScienceFair2025 folder into the Terminal window** and press Enter.
2. Run the file
#### (Images can be flipped and rotated. This is *normal* and not an issue)
```python3 Scripts/datacheck.py```
3. **Wait for the dataset to load**
4. Follow in-terminal instructions
###