CSC426 - Machine Learning :: Dr. Bloodgood

- Colin Friedlander, Tyler Elvis, Joey Crane, Mina Guglietta

---

OBJECTIVE

Our objective is to create a machine, learning algorithm for tic-tac-toe that never loses That never loses and finds the most optimal move to win given the board state

---

Commit Convention

We use a Conventional Commits–style format so the git history is easy to scan.

<type>(<scope>): <subject>

Type	Purpose
feat	New feature
fix	Bug fix
docs	Documentation only e
refactor	Code change that doesn't alter behavior
chore	Maintenance (deps, formatting, repo hygiene)
ci	CI/CD and workflow changes

Examples:

• feat(ui): add GlassIconButton
• fix(api): handle null patientId
• ci: install pnpm via npm
• docs(readme): add commit convention section
• chore: bump deps
• feat(api)!: rename /v1/messages to /v2/messages ← breaking change

Keep subjects short (< 72 chars), imperative mood ("add", "fix", "remove").

---

Command Line

#Adding matplotlib
module load python/3.10.11

python -m pip install matplotlib

# Running The Main 
cd to ML_Project_One

python3 -m src.main

---

File Structure:

• ML_Project_One:

  • src:

    • main.py
    • game_rules.py
    • model.py

  • graphs:

    • graphs.py
    • figures

  • logs:

    • log_utils.py
    • log.txt

  • csv:

    • model_weights.csv
    • wins.csv
    • losses.csv
    • draws.csv

  • README.MD

---

What's In What?

- src

• main.py, responsible for running the whole program and model training

• game_rules.py responsible for handling all things board related

• model.py responsible for handling all functions required to create the learning model

• graphs

  • graphs.py responsible for creating all graphs

- logs

• log_utils.py responsible for handling loging information like errors or logging data into a csv

- csv

• model_weights.csv responsible for holding all weights changes of the model
• wins.csv, losses.csv, draws.csv responsible for holding info on wins, losses, draws

- README.MD responsible for the readme