https://github.com/rkarahul/ok.win-big-small-predictor
Predict the next “Big” or “Small” outcome on the OK.Win lottery-style game using OCR + time-series features + ML.
https://github.com/rkarahul/ok.win-big-small-predictor
joblib numpy opencv-python paddleocr paddlepaddle paddlepaddle-gpu pandas python sckit-learn
Last synced: 2 months ago
JSON representation
Predict the next “Big” or “Small” outcome on the OK.Win lottery-style game using OCR + time-series features + ML.
- Host: GitHub
- URL: https://github.com/rkarahul/ok.win-big-small-predictor
- Owner: rkarahul
- Created: 2025-10-30T11:52:27.000Z (8 months ago)
- Default Branch: main
- Last Pushed: 2025-10-30T11:56:49.000Z (8 months ago)
- Last Synced: 2025-10-30T13:36:12.304Z (8 months ago)
- Topics: joblib, numpy, opencv-python, paddleocr, paddlepaddle, paddlepaddle-gpu, pandas, python, sckit-learn
- Language: Python
- Homepage:
- Size: 429 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# OK.Win Big/Small Predictor
**Predict the next “Big” or “Small” outcome on the OK.Win lottery-style game using OCR + time-series features + ML.**
---
## 📌 Project Overview
1. **OCR Extraction** – PaddleOCR reads game screenshots → structured CSV.
2. **Feature Engineering** – Parse the 17-digit `Period` into year/month/day/hour/minute/second/serial.
3. **Modeling** – 7 classic classifiers; best model (usually Random Forest) is saved.
4. **Inference** – Load `best_model.pkl` + `scaler.pkl` to predict live screenshots.
---
## 🚀 Quick Start
```bash
# 1. Clone
git clone https://github.com/rkarahul/OK.Win-Big-Small-Predictor.git
cd okwin-bigsmall-predictor
# 2. Install
pip install -r requirements.txt
# 3. Run pipeline (images → model)
python src/01_ocr_extract.py # needs folder ./images/
python src/02_format_data.py
python src/03_train_model.py