How huseinzol05/stock-prediction-models works
Stock-Prediction-Models is a collection of stock-forecasting notebooks, trading agents, simulations, data explorations, and a browser-based TensorFlow.js demo—not one unified runtime. The README describes independent model and analysis capabilities; the graph preserves those branches without implying a shared processing pipeline. Sampled source confirms the real-time Flask service’s trade/reset interface and the DNC’s controller-to-memory structure. Most notebook implementations were not sampled, so their internal relationships are omitted.
Main entry points: stacking Stacking models · app.py Trade and reset API · access.py Memory access · addressing.py Memory addressing
Components 20
Forecasting models
- Deep-learning models
deep-learning - DNC core
deep-learning/dnc.py - Memory access
deep-learning/access.py - Memory addressing
deep-learning/addressing.py - DNC utilities
deep-learning/util.py - Stacking models
stacking - Stacking autoencoder
stacking/autoencoder.py - Stacking RNN
stacking/model.py
Trading agents
- Trading strategies
agent - Evolution strategies
free-agent
Analysis and simulation
- Market datasets
dataset - Market simulations
simulation - Data explorations
misc
Real-time service
- API client
- Trade and reset API
realtime-agent/app.py - Evolution trading agent
realtime-agent/app.py
Browser demo
- Forecasting interface
stock-forecasting-js/init.js - TensorFlow.js utilities
stock-forecasting-js/js/utils.js - Google sample data
stock-forecasting-js/data/google.js
Other components
- Engineer or analyst
Connections 15
- Engineer or analyst → Deep-learning models: explores forecasts
README.md - Engineer or analyst → Stacking models: explores ensembles
README.md - Engineer or analyst → Trading strategies: runs strategies
README.md - Engineer or analyst → Evolution strategies: explores agents
README.md - Engineer or analyst → Market simulations: runs simulations
README.md - Engineer or analyst → Data explorations: studies data
README.md - Engineer or analyst → Forecasting interface: uses demo
README.md - API client → Trade and reset API: requests trade
realtime-agent/app.py - Trade and reset API → Evolution trading agent: dispatches trade
realtime-agent/app.py - Trade and reset API → Evolution trading agent: resets capital
realtime-agent/app.py - DNC core → Memory access: calls memory
deep-learning/dnc.py - Memory access → Memory addressing: uses addressing
deep-learning/access.py - Memory addressing → DNC utilities: uses utilities
deep-learning/addressing.py - Stacking models → Stacking autoencoder: reduces dimensions
README.md - Stacking models → Stacking RNN: includes recurrent model
README.md