JonusNattapong/xauusd-dataset

XAUUSD Enhanced ML Dataset

Comprehensive machine learning dataset for XAUUSD (Gold vs US Dollar) price prediction with 172 advanced features.

Dataset Description

This dataset contains cleaned and processed XAUUSD price data optimized for machine learning applications. It includes 172 features covering technical indicators, economic variables, statistical measures, and temporal features.

Key Features:

• Time Period: 2023-2025 (708 observations)
• Features: 172 advanced technical and economic indicators
• Data Quality: Cleaned, no missing values, processed for ML
• Target Variables: Binary classification for price direction prediction
• ML Performance: 47.3% directional accuracy with ensemble models

Feature Categories:

Technical Indicators (85+ features):

• Volume Indicators: ADI, OBV, CMF, FI, EM, SMA_EM, VPT, VWAP, MFI, NVI
• Volatility Measures: Bollinger Bands, Keltner Channels, Donchian Channels, ATR, UI
• Trend Indicators: MACD, SMA, EMA, Vortex, TRIX, Mass Index, DPO, KST, Ichimoku, STC, ADX, CCI, Aroon, Parabolic SAR
• Momentum Indicators: RSI, Stochastic RSI, TSI, Ultimate Oscillator, Stochastic, Williams %R, AO, ROC, PPO, PVO, KAMA
• Other: DR, DLR, CR

Economic & Market Data:

• Currency: DXY (US Dollar Index)
• Bonds: US 10Y Treasury Yield
• Commodities: WTI Oil, Silver, Copper, BTC
• Ratios: Gold/Silver, Gold/Oil, Gold/Copper, Gold/BTC

Statistical & Temporal Features:

• Price Changes: 1d, 5d, 20d percentage changes
• Volatility: Rolling volatility (5d, 20d), Realized volatility
• Rolling Statistics: Mean, Std, Skew, Kurtosis (5, 10, 20, 50 periods)
• Lagged Features: Price and return lags (1, 2, 3, 5, 10, 20 days)
• Risk Metrics: VaR (95%, 99%), CVaR, Sharpe ratio, Max drawdown
• Seasonal: Day of week, month, quarter, sine/cosine transformations

Machine Learning Target:

• Binary Classification: Price direction prediction (up/down)
• Directional Accuracy: 47.3% achieved with ensemble models
• Cross-validation: Time series split with expanding window

Usage

Load with Pandas (Recommended):

import pandas as pd

# Load the enhanced ML dataset
df = pd.read_csv("https://huggingface.co/datasets/JonusNattapong/xauusd-dataset/resolve/main/XAUUSD_enhanced_ml_dataset_clean.csv")
print(f"Dataset shape: {df.shape}")
print(f"Features: {len(df.columns)}")

Load with Hugging Face Datasets:

Note: Due to multiple CSV files with different schemas in this repository, the HF datasets library may encounter compatibility issues. Direct CSV loading (above) is recommended for best results.

If you prefer to use the datasets library, you can load the CSV directly:

from datasets import load_dataset

# Load the specific CSV file
dataset = load_dataset('csv', data_files="https://huggingface.co/datasets/JonusNattapong/xauusd-dataset/resolve/main/XAUUSD_enhanced_ml_dataset_clean.csv")
print(dataset['train'].column_names)
print(dataset['train'][0])

Example ML Workflow:

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score

# Load data
df = pd.read_csv("https://huggingface.co/datasets/JonusNattapong/xauusd-dataset/resolve/main/XAUUSD_enhanced_ml_dataset_clean.csv")

# Prepare features and target
feature_cols = [col for col in df.columns if col not in ['Date', 'Target_1d', 'Target_5d']]
X = df[feature_cols]
y = df['Target_1d']

# Split data (time series aware)
split_idx = int(len(df) * 0.8)
X_train, X_test = X[:split_idx], X[split_idx:]
y_train, y_test = y[:split_idx], y[split_idx:]

# Train model
model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

# Evaluate
y_pred = model.predict(X_test)
accuracy = accuracy_score(y_test, y_pred)
print(f"Accuracy: {accuracy:.3f}")

Performance Benchmark

The dataset was used to train ensemble models achieving:

• Directional Accuracy: 47.3%
• Top Features: Rolling volatility, momentum indicators, RSI, MACD
• Cross-validation: Time series split with expanding window

Citation

If you use this dataset in your research, please cite:

@misc{{tapachoom2025xauusd,
  title={{XAUUSD Enhanced ML Dataset}},
  author={{Tapachoom, Nattapong}},
  year={{2025}},
  publisher={{Hugging Face}},
  url={{https://huggingface.co/datasets/JonusNattapong/xauusd-dataset}}
}}

License

This dataset is available under the MIT License for educational and research purposes.