AlphaTrade

🧠 Architecture

Ticker · RELIANCE.NS
Model · LSTM 64→Dense 32
Lookback · 5-Day Window
Features · 16 Hybrid
NLP · VADER + FAISS RAG

Reliance Industries (RELIANCE.NS)

AI Stock Market Prediction System

⚡ Model Online🕒 6 Months
IEEE Research BenchmarkDeterministic Seed (v2.0)

Experimental Results & Ablation Study

Systematic empirical evaluation comparing 4 architectural variants across 16 technical, sentiment, and event features.

Peak Recall
95.65%

Hybrid RAG-LSTM Model

Peak ROC-AUC
0.9519

Event-Augmented Variant

Inference Latency
2.16 ms

Production Streaming Ready

Evaluated Variants
4 Ablations

8 to 16 Features

Model Variant Benchmark Matrix

Empirical testing with 80/20 deterministic train-test split on daily OHLCV and event news series

Loss: Binary Focal (γ=2.0)
Model VariantFeaturesAccuracyPrecisionRecallF1-ScoreROC-AUCLatency

Visual Metric Comparison

Measures sensitivity to price directional changes. Crucial in trading to avoid missing catastrophic downturns.

Confusion Matrix

Selected

Click any model row in the matrix table above to inspect its classification error distribution.

Quantitative Insight: The proposed RAG-LSTM exhibits the lowest False Negative count (1), ensuring that unexpected upside or catalyst surges are never left uncaptured.
Why Recall is Priority #1

In quantitative trading, a False Negative represents holding through an unexpected crash catalyst. Our 95.65% Recall significantly outperforms pure technical models.

Ultra-Low Latency (2.16 ms)

With FAISS C++ vector indexing and compact 384-dim embeddings, inference runs in just 2.16 milliseconds per sample on standard CPUs without requiring high-end GPUs.

Binary Focal Loss Advantage

Standard Cross-Entropy fails on choppy sideways markets. Dynamic Focal Loss (γ=2.0, α=0.25) penalizes hard-to-classify edge days, preventing model collapse.