Time-Series Forecasting for Market Trends (Prophet/LSTM)

[Cedarich]

Labels: Backend, AI/ML, Analytics
Complexity: High (200 points) 📈

Description

Implement a predictive analytics layer to forecast market trends (Bullish/Bearish) for the next 24-48 hours based on historical sentiment and volume data.

Requirements / Context

• Use libraries like prophet (Meta) or tensorflow/pytorch for simple LSTM.
• Predicts "Sentiment Velocity" (how fast mood is changing).

Acceptance Criteria

• src/analytics/forecaster.py implemented.
• Training script to build model from analytics.jsonl history.
• API endpoint GET /analytics/forecast returning predicted scores.

---

[rosemary21]

@rosemary21 has applied to work on this issue as part of the Stellar Wave Program's 3rd wave.

I’d like to work on this issue because it sits at the intersection of backend engineering, analytics, and machine learning—areas I’m deeply interested in. Building a predictive layer that translates historical sentiment and volume data into actionable short-term forecasts (Bullish/Bearish trends) is both technically challenging and highly impactful. It adds real intelligence to the system rather than just reporting past data.

I have experience working with time-series forecasting and ML pipelines, including:

📊 Time-Series & Forecasting
Built forecasting models using Facebook Prophet for trend and seasonality analysis
Worked with sequence models (LSTM/GRU) in TensorFlow and PyTorch for temporal pattern prediction
Experience handling rolling windows, lag features, and feature engineering for time-series datasets
🤖 Machine Learning Pipelines
Created end-to-end training pipelines (data ingestion → preprocessing → training → evaluation)
Worked with JSONL datasets similar to analytics.jsonl for incremental model training
Implemented model versioning and reproducible training scripts
⚙️ Backend Integration
Integrated ML models into APIs (e.g., REST endpoints serving predictions)
Designed endpoints similar to /analytics/forecast that return real-time inference results
Experience optimizing inference for low-latency responses

ℹ️ Repo Maintainers: To accept this application, review their application or assign @rosemary21 to this issue.

[drips-wave]

Congratulations, @rosemary21! 🎉 Your application was accepted by the repo's maintainers, and the issue is due on March 30, 2026.

🧑‍💻 @rosemary21: Please resolve the issue such that the repo's maintainers have enough time to review your contribution before the due date. You'll earn Points for completing the issue on-time, which will make you eligible for a share of the Stellar Wave Program's reward pool.

Warning

When opening a PR, please link it to this issue to ensure it gets tracked accurately. Points are awarded when this issue is marked as completed by the maintainer.

🤠 Repo maintainers: Please keep an eye on the contributor's progress and review their work before the due date. You can manage this issue, including adjusting its complexity and points, here.

🌊 Happy Wave 🌊

[drips-wave]

This issue has been marked as completed by a Drips Wave moderator for @rosemary21 as part of the Stellar Wave Program's 3rd Wave 🥳

Moderator's reason: [Auto-assessed] The contributor successfully implemented the predictive analytics layer, including the forecaster class, a training script, and the required API endpoint. The PR was merged by maintainers and fully addresses the issue's acceptance criteria.

😎 @rosemary21: You earned 200 Points for completing this issue! After the current Wave ends, you'll be eligible for a percentage of the Wave's reward pool based on the percentage of total points you've earned. Learn more here.

🧑‍💻 Repo maintainers: How'd the contributor do? Leave a review to share your experience working with them.

ℹ️ Note: This issue was marked complete via moderation. Points were issued even though the GitHub issue remains open.