nse-lstm-model / model_card.json
thoutam's picture
Update README with YAML metadata and enhance model card
95b67c4 verified
{
"model_name": "nse-lstm-model",
"model_type": "LSTM Neural Network",
"task": "Stock Price Prediction",
"dataset": "NSE Bhavcopy (2004-2025)",
"license": "MIT",
"author": "thoutam",
"description": "A comprehensive LSTM model trained on 6.8 million records across 3,622 symbols from the National Stock Exchange (NSE) of India for stock price prediction using technical indicators and historical patterns.",
"metrics": {
"training_mae": 0.0216,
"validation_mae": 0.0217,
"training_mse": 0.0005,
"validation_mse": 0.0005
},
"architecture": {
"input_shape": [5, 25],
"output_shape": [1],
"total_params": 60225,
"trainable_params": 60225,
"non_trainable_params": 0,
"layers": [
"LSTM(64) + Dropout(0.2)",
"LSTM(64) + Dropout(0.2)",
"Dense(32)",
"Dense(16)",
"Dense(1)"
]
},
"features": [
"OHLCV data",
"Moving Averages (5, 10, 20, 50 day)",
"Bollinger Bands (20 day)",
"RSI (14 day)",
"MACD",
"Volume indicators (OBV, VPT)"
],
"training_data": {
"total_records": 6795445,
"symbols": 3622,
"date_range": "2004-2025",
"frequency": "daily",
"source": "NSE Bhavcopy"
},
"usage": {
"input_format": "Normalized features with shape (batch_size, 5, 25)",
"output_format": "Single prediction value for next day's price change",
"preprocessing": "Requires StandardScaler normalization using provided scaler.pkl"
},
"limitations": [
"Trained on Indian market data only",
"Requires 5 days of historical data",
"Predictions are based on technical indicators only"
],
"tags": [
"finance",
"stock-market",
"lstm",
"time-series",
"prediction",
"nse",
"india",
"technical-analysis"
]
}