# Attention Score Model Training Report

**Generated:** 2026-01-06 23:46:55

## Overview

The attention model learns to identify high-flow market moments using volume, volatility, and money flow indicators - WITHOUT hardcoding specific trading hours or sessions.

## Configuration

- **Symbols:** XAUUSD, EURUSD
- **Timeframes:** 5m, 15m
- **Training Data Cutoff:** 2024-03-01
- **Training Years:** 5.0
- **Holdout Years:** 1.0

### Model Parameters

| Parameter | Value |
|-----------|-------|
| Factor Window | 200 |
| Horizon Bars | 3 |
| Low Flow Threshold | 1.0 |
| High Flow Threshold | 2.0 |

### Features Used (9 total)

| Feature | Description |
|---------|-------------|
| volume_ratio | volume / rolling_median(volume, 20) |
| volume_z | z-score of volume over 20 periods |
| ATR | Average True Range (14 periods) |
| ATR_ratio | ATR / rolling_median(ATR, 50) |
| CMF | Chaikin Money Flow (20 periods) |
| MFI | Money Flow Index (14 periods) |
| OBV_delta | diff(OBV) / rolling_std(OBV, 20) |
| BB_width | (BB_upper - BB_lower) / close |
| displacement | (close - open) / ATR |

## Training Results

| Model | Symbol | TF | Reg MAE | Reg R2 | Clf Acc | Clf F1 | N Train | High Flow % |
|-------|--------|-----|---------|--------|---------|--------|---------|-------------|
| XAUUSD_5m_attention | XAUUSD | 5m | 0.8528 | 0.1914 | 61.44% | 57.96% | 288386 | 23.1% |
| XAUUSD_15m_attention | XAUUSD | 15m | 0.8564 | 0.1250 | 59.39% | 54.70% | 96801 | 25.8% |
| EURUSD_5m_attention | EURUSD | 5m | 0.6678 | 0.1569 | 54.07% | 49.84% | 312891 | 34.3% |
| EURUSD_15m_attention | EURUSD | 15m | 0.6405 | 0.2193 | 60.70% | 57.20% | 104659 | 36.3% |


## Class Distribution (Holdout Set)

| Model | Low Flow | Medium Flow | High Flow |
|-------|----------|-------------|-----------|
| XAUUSD_5m_attention | 265 (0.4%) | 53705 (76.5%) | 16238 (23.1%) |
| XAUUSD_15m_attention | 0 (0.0%) | 17566 (74.2%) | 6106 (25.8%) |
| EURUSD_5m_attention | 2380 (3.2%) | 46893 (62.5%) | 25781 (34.3%) |
| EURUSD_15m_attention | 443 (1.8%) | 15629 (62.0%) | 9143 (36.3%) |


## Feature Importance

### XAUUSD_5m_attention

| Rank | Feature | Combined Importance |
|------|---------|--------------------|
| 1 | ATR_ratio | 0.4240 |
| 2 | BB_width | 0.1601 |
| 3 | ATR | 0.1229 |
| 4 | CMF | 0.1164 |
| 5 | volume_ratio | 0.0639 |
| 6 | volume_z | 0.0399 |
| 7 | displacement | 0.0331 |
| 8 | OBV_delta | 0.0213 |
| 9 | MFI | 0.0184 |

### XAUUSD_15m_attention

| Rank | Feature | Combined Importance |
|------|---------|--------------------|
| 1 | ATR_ratio | 0.3364 |
| 2 | volume_ratio | 0.1779 |
| 3 | BB_width | 0.1414 |
| 4 | volume_z | 0.1034 |
| 5 | displacement | 0.0743 |
| 6 | ATR | 0.0651 |
| 7 | OBV_delta | 0.0441 |
| 8 | CMF | 0.0331 |
| 9 | MFI | 0.0243 |

### EURUSD_5m_attention

| Rank | Feature | Combined Importance |
|------|---------|--------------------|
| 1 | ATR_ratio | 0.3577 |
| 2 | BB_width | 0.2217 |
| 3 | ATR | 0.1566 |
| 4 | volume_ratio | 0.0765 |
| 5 | CMF | 0.0569 |
| 6 | volume_z | 0.0536 |
| 7 | displacement | 0.0315 |
| 8 | OBV_delta | 0.0264 |
| 9 | MFI | 0.0191 |

### EURUSD_15m_attention

| Rank | Feature | Combined Importance |
|------|---------|--------------------|
| 1 | ATR_ratio | 0.5007 |
| 2 | volume_ratio | 0.1497 |
| 3 | volume_z | 0.1129 |
| 4 | ATR | 0.0990 |
| 5 | BB_width | 0.0396 |
| 6 | displacement | 0.0284 |
| 7 | CMF | 0.0254 |
| 8 | OBV_delta | 0.0245 |
| 9 | MFI | 0.0198 |



## Interpretation

### Attention Score (Regression)

- **< 1.0**: Low flow period - below average market movement expected
- **1.0 - 2.0**: Medium flow period - average market conditions
- **> 2.0**: High flow period - above average movement expected (best trading opportunities)

### Flow Class (Classification)

- **0 (low_flow)**: move_multiplier < 1.0
- **1 (medium_flow)**: 1.0 <= move_multiplier < 2.0
- **2 (high_flow)**: move_multiplier >= 2.0

## Trading Recommendations

1. **Filter by attention_score**: Only trade when attention_score > 1.0
2. **Adjust position sizing**: Increase size when attention_score > 2.0
3. **Combine with base models**: Use attention_score as feature #51 in prediction models
4. **Time-agnostic**: The model identifies flow without hardcoded sessions

## Usage Example

```python
from training.attention_trainer import AttentionModelTrainer

# Load trained models
trainer = AttentionModelTrainer.load('models/attention/')

# Get attention score for new OHLCV data
attention = trainer.get_attention_score(df_ohlcv, 'XAUUSD', '5m')

# Filter trades
mask_trade = attention > 1.0  # Only trade in medium/high flow

# Or use as feature in base models
df['attention_score'] = attention
```

## Files Generated

- `models/attention/{symbol}_{timeframe}_attention/` - Model directories
- `models/attention/trainer_metadata.joblib` - Trainer configuration
- `models/attention/training_summary.csv` - Summary metrics

---
*Report generated by Attention Model Training Pipeline*