NLP-Commodity-Signaller kshitijshresth · COMPLETED
Sentiment analysis pipeline for commodity trading using DistilBERT, MiniLM, and GloVe-LSTM models. Processes real-time NewsData.io, MarketAux, Reddit data to predict 1-5 sentiment scores, mapping to Buy/Sell/Hold signals. Features oversampling, class-weighted loss, and backtesting with regression/classification metrics.
github.com/kshitijshresth/NLP-Commodity-Signaller · ★ 0 · Forks 0 · Size 2.4 MB
SUMMARY
Technologies 2
Scored 2
Observed 2
Practices 6
Evidence 2
Skips 0
COVERAGE
Analyzed 8 files · 6 commits · 0 API calls
TECHNOLOGIES & DEPTH
Markdown LANGUAGE Depth 70
1 files · PRODUCTION
MongoDB DATABASE Depth 61
1 files · CONFIGURATION
PRACTICES
documentation · observedautomated_tests · absentcontinuous_integration · absentcontainerization · absentlinting · absentformatting · absent
ACTIVITY & OWNERSHIP
First commit 2025-06-08
Last commit 2025-06-08
Active months 1
Commits 6