A native desktop (non-browser) machine learning application.
Features:
- Classification
- Regression
- Clustering
- Anomaly Detection
- Dimensionality Reduction
🛠️ Running from Source (Development Mode)
If you want to run the app from source code instead of the EXE:
conda create -n ml-desktop python=3.10 -y conda activate ml-desktop pip install -r requirements.txt python main.py
🏗️ Build Windows Executable (EXE)
To build a standalone Windows app:
pip install pyinstaller pyinstaller --noconsole --onefile --name "ML-Trainer" main.py
The executable will be created in:
dist/ML-Trainer.exe
Python 3.10 is required
First launch may take a few seconds (PyInstaller extraction)
Do NOT commit dist/ or build/ folders to git
Upload EXE to GitHub Releases, not source tree
If antivirus flags the EXE → it’s a PyInstaller false positive
📦 Recommended .gitignore
pycache/ *.pyc
.env .venv
build/ dist/ *.exe *.spec
.DS_Store Thumbs.db
🚀 Versioning
v1.0.0 → First stable desktop release
Source = GitHub repo
Binary = GitHub Releases