Upload folder using huggingface_hub

frontend/app/components/dashboard/StockPredictions.tsx

@@ -89,7 +89,7 @@ const StockPredictions = () => {
         <div className="flex items-center justify-between mb-4">
           <div className="flex items-center gap-2">
             <Activity className="w-5 h-5 text-success" />
-            <h2 className="text-lg font-bold">STOCK PREDICTIONS - ML MODELS</h2>
+            <h2 className="text-lg font-bold">CSE STOCK PREDICTIONS 🇱🇰</h2>
           </div>
           <div className="flex items-center gap-2">
             <button
@@ -165,7 +165,7 @@ const StockPredictions = () => {
                   </div>
                   <div className="text-right">
                     <div className="font-mono text-lg">
-                      ${stock.predicted_price?.toFixed(2) || '---'}
+                      LKR {stock.predicted_price?.toFixed(2) || '---'}
                     </div>
                     <div className={`text-sm font-mono ${stock.expected_change_pct >= 0 ? 'text-green-400' : 'text-red-400'}`}>
                       {stock.expected_change_pct >= 0 ? '+' : ''}{stock.expected_change_pct?.toFixed(2) || '0.00'}%

main.py

@@ -15,7 +15,7 @@ from pydantic import BaseModel
 from typing import Dict, Any, List, Set, Optional
 import asyncio
 import json
-from datetime import datetime
+from datetime import datetime, timedelta
 import sys
 import os
 import logging

models/stock-price-prediction/README.md

@@ -0,0 +1,85 @@
+# Stock Price Prediction Module 🇱🇰
+
+BiLSTM-based stock price prediction for **10 Sri Lankan CSE stocks**.
+
+## Stocks Covered
+
+| Symbol | Company | Sector |
+|--------|---------|--------|
+| COMB | Commercial Bank of Ceylon PLC | Banking |
+| JKH | John Keells Holdings PLC | Diversified Holdings |
+| SAMP | Sampath Bank PLC | Banking |
+| HNB | Hatton National Bank PLC | Banking |
+| DIAL | Dialog Axiata PLC | Telecommunications |
+| CTC | Ceylon Tobacco Company PLC | Consumer Goods |
+| NEST | Nestle Lanka PLC | Consumer Goods |
+| CARG | Cargills Ceylon PLC | Retail |
+| HNBA | HNB Assurance PLC | Insurance |
+| CARS | Carson Cumberbatch PLC | Diversified Holdings |
+
+## ⚠️ Important Note
+
+**Yahoo Finance does NOT support CSE (Colombo Stock Exchange) tickers directly.**
+
+The module uses fallback predictions with simulated market data. For real CSE data, integrate with:
+- CSE official API
+- Bloomberg Terminal
+- Reuters/Refinitiv
+
+## Architecture
+
+- **Model**: Bidirectional LSTM (BiLSTM)
+- **Epochs**: 10 (configurable)
+- **Sequence Length**: 60 days
+- **Features**: Close price, technical indicators
+- **Tracking**: MLflow + DagsHub
+
+## Quick Start
+
+```bash
+# Train all 10 stocks
+cd models/stock-price-prediction
+python main.py
+
+# Test predictor
+python src/components/predictor.py
+```
+
+## API Endpoints
+
+| Endpoint | Description |
+|----------|-------------|
+| `GET /api/stocks/predictions` | All 10 stock predictions |
+| `GET /api/stocks/predictions/{symbol}` | Single stock (COMB, JKH, etc.) |
+| `GET /api/stocks/model/status` | Model training status |
+
+## Output
+
+Predictions include:
+- Current price (LKR)
+- Predicted next-day price
+- Expected change %
+- Trend (bullish/bearish/neutral)
+- Confidence score
+
+## Directory Structure
+
+```
+stock-price-prediction/
+├── main.py                 # Multi-stock training entry
+├── src/
+│   ├── components/
+│   │   ├── data_ingestion.py
+│   │   ├── data_validation.py
+│   │   ├── data_transformation.py
+│   │   ├── model_trainer.py
+│   │   └── predictor.py    # Inference API
+│   └── constants/
+│       └── training_pipeline/
+├── Artifacts/              # Trained models
+└── output/predictions/     # JSON predictions
+```
+
+## Airflow DAG
+
+Schedule: **4:15 AM IST daily** (via centralized `airflow/dags/stock_prediction_dag.py`)

models/stock-price-prediction/src/components/predictor.py

@@ -97,17 +97,25 @@ class StockPredictor:
         """Generate a fallback prediction when model is not available."""
         stock_info = STOCKS_TO_TRAIN.get(stock_code, {"name": stock_code, "sector": "Unknown"})
         
-        # Simulated current price (would normally fetch from Yahoo Finance)
-        np.random.seed(hash(stock_code) % 2**31)
-        base_prices = {
-            "AAPL": 175.0, "GOOGL": 140.0, "MSFT": 380.0, "AMZN": 180.0,
-            "META": 500.0, "NVDA": 480.0, "TSLA": 250.0, "JPM": 190.0,
-            "V": 275.0, "JNJ": 155.0
+        # Realistic CSE stock prices in LKR (Sri Lankan Rupees)
+        # Based on typical market cap leaders on CSE
+        np.random.seed(hash(stock_code + datetime.now().strftime("%Y%m%d")) % 2**31)
+        base_prices_lkr = {
+            "COMB": 95.0,   # Commercial Bank ~95 LKR
+            "JKH": 175.0,   # John Keells Holdings ~175 LKR
+            "SAMP": 68.0,   # Sampath Bank ~68 LKR
+            "HNB": 155.0,   # Hatton National Bank ~155 LKR
+            "DIAL": 12.0,   # Dialog Axiata ~12 LKR
+            "CTC": 1100.0,  # Ceylon Tobacco ~1100 LKR
+            "NEST": 1450.0, # Nestle Lanka ~1450 LKR
+            "CARG": 215.0,  # Cargills Ceylon ~215 LKR
+            "HNBA": 42.0,   # HNB Assurance ~42 LKR
+            "CARS": 285.0,  # Carson Cumberbatch ~285 LKR
         }
-        current_price = base_prices.get(stock_code, 100.0) * (1 + np.random.uniform(-0.05, 0.05))
+        current_price = base_prices_lkr.get(stock_code, 100.0) * (1 + np.random.uniform(-0.03, 0.03))
         
         # Generate prediction with slight randomized movement
-        change_pct = np.random.normal(0.1, 1.0)  # Mean +0.1%, std 1%
+        change_pct = np.random.normal(0.15, 1.5)  # Mean +0.15%, std 1.5%
         predicted_price = current_price * (1 + change_pct / 100)
         
         # Determine trend
@@ -125,6 +133,8 @@ class StockPredictor:
             "symbol": stock_code,
             "name": stock_info.get("name", stock_code),
             "sector": stock_info.get("sector", "Unknown"),
+            "exchange": stock_info.get("exchange", "CSE"),
+            "currency": "LKR",
             "current_price": round(current_price, 2),
             "predicted_price": round(predicted_price, 2),
             "expected_change": round(predicted_price - current_price, 2),
@@ -134,7 +144,7 @@ class StockPredictor:
             "confidence": round(np.random.uniform(0.65, 0.85), 2),
             "model_architecture": "BiLSTM",
             "is_fallback": True,
-            "note": "Using fallback predictions - train models for accurate forecasts"
+            "note": "CSE data via fallback - Yahoo Finance doesn't support CSE tickers"
         }
     
     def predict_stock(self, stock_code: str) -> Dict[str, Any]:

models/stock-price-prediction/src/constants/training_pipeline/__init__.py

@@ -5,64 +5,75 @@ import numpy as np
 Defining common constant variable for training pipeline
 """
 
-# Stocks available on Yahoo Finance for training
-# NOTE: CSE (Sri Lanka) tickers are NOT available on Yahoo Finance
-# Using globally available tickers instead
+# Top 10 Sri Lankan Stocks by market cap (CSE - Colombo Stock Exchange)
+# NOTE: Yahoo Finance does NOT support CSE tickers directly.
+# These stocks will use fallback predictions with simulated market data.
+# For real CSE data, integrate with CSE API or data providers like Bloomberg.
 STOCKS_TO_TRAIN = {
-    "AAPL": {
-        "yahoo_symbol": "AAPL",
-        "name": "Apple Inc.",
-        "sector": "Technology"
+    "COMB": {
+        "yahoo_symbol": "COMB.N0000",  # Commercial Bank of Ceylon
+        "name": "Commercial Bank of Ceylon PLC",
+        "sector": "Banking",
+        "exchange": "CSE"
     },
-    "GOOGL": {
-        "yahoo_symbol": "GOOGL",
-        "name": "Alphabet Inc.",
-        "sector": "Technology"
+    "JKH": {
+        "yahoo_symbol": "JKH.N0000",  # John Keells Holdings
+        "name": "John Keells Holdings PLC",
+        "sector": "Diversified Holdings",
+        "exchange": "CSE"
     },
-    "MSFT": {
-        "yahoo_symbol": "MSFT",
-        "name": "Microsoft Corporation",
-        "sector": "Technology"
+    "SAMP": {
+        "yahoo_symbol": "SAMP.N0000",  # Sampath Bank
+        "name": "Sampath Bank PLC",
+        "sector": "Banking",
+        "exchange": "CSE"
     },
-    "AMZN": {
-        "yahoo_symbol": "AMZN",
-        "name": "Amazon.com Inc.",
-        "sector": "Consumer Discretionary"
+    "HNB": {
+        "yahoo_symbol": "HNB.N0000",  # Hatton National Bank
+        "name": "Hatton National Bank PLC",
+        "sector": "Banking",
+        "exchange": "CSE"
     },
-    "META": {
-        "yahoo_symbol": "META",
-        "name": "Meta Platforms Inc.",
-        "sector": "Technology"
+    "DIAL": {
+        "yahoo_symbol": "DIAL.N0000",  # Dialog Axiata
+        "name": "Dialog Axiata PLC",
+        "sector": "Telecommunications",
+        "exchange": "CSE"
     },
-    "NVDA": {
-        "yahoo_symbol": "NVDA",
-        "name": "NVIDIA Corporation",
-        "sector": "Technology"
+    "CTC": {
+        "yahoo_symbol": "CTC.N0000",  # Ceylon Tobacco
+        "name": "Ceylon Tobacco Company PLC",
+        "sector": "Consumer Goods",
+        "exchange": "CSE"
     },
-    "TSLA": {
-        "yahoo_symbol": "TSLA",
-        "name": "Tesla Inc.",
-        "sector": "Automotive"
+    "NEST": {
+        "yahoo_symbol": "NEST.N0000",  # Nestle Lanka
+        "name": "Nestle Lanka PLC",
+        "sector": "Consumer Goods",
+        "exchange": "CSE"
     },
-    "JPM": {
-        "yahoo_symbol": "JPM",
-        "name": "JPMorgan Chase & Co.",
-        "sector": "Financial Services"
+    "CARG": {
+        "yahoo_symbol": "CARG.N0000",  # Cargills Ceylon
+        "name": "Cargills Ceylon PLC",
+        "sector": "Retail",
+        "exchange": "CSE"
     },
-    "V": {
-        "yahoo_symbol": "V",
-        "name": "Visa Inc.",
-        "sector": "Financial Services"
+    "HNBA": {
+        "yahoo_symbol": "HNBA.N0000",  # HNB Assurance
+        "name": "HNB Assurance PLC",
+        "sector": "Insurance",
+        "exchange": "CSE"
     },
-    "JNJ": {
-        "yahoo_symbol": "JNJ",
-        "name": "Johnson & Johnson",
-        "sector": "Healthcare"
+    "CARS": {
+        "yahoo_symbol": "CARS.N0000",  # Carson Cumberbatch
+        "name": "Carson Cumberbatch PLC",
+        "sector": "Diversified Holdings",
+        "exchange": "CSE"
     }
 }
 
 # Default stock for single-stock training mode
-DEFAULT_STOCK = "AAPL"
+DEFAULT_STOCK = "COMB"
 
 # Legacy alias for backward compatibility
 SRI_LANKA_STOCKS = STOCKS_TO_TRAIN