[help wanted] Improve pump scanner detection accuracy

[stefanoviana]

The pump scanner currently uses volume spike + RSI + 24h change to detect pumps.

Ideas to improve:

• Order flow imbalance (bid vs ask volume)
• Social media sentiment (Twitter/Telegram mentions)
• Exchange listing detection
• Whale wallet tracking
• Machine learning on historical pump data

Current params:

• Volume spike: 5x average
• RSI: < 85
• Cooldown: 30min after close
• Max 2 attempts per coin per day

Share your ideas or submit a PR!

Difficulty: Advanced

[github-actions]

Thanks for opening an issue!

Quick setup:

pip install -r requirements.txt
cp .env.example .env
python deepalpha.py

Need help? Join our Discord or check the Dashboard.

Want Pro features? DeepAlpha Pro includes 72 features, ensemble models, cloud dashboard, and priority support.

We'll review your issue soon!