Tipo de estrategia
Algoritmos de trading con aprendizaje automático
Modelos ML de código abierto para finanzas cuantitativas
Los algoritmos de trading con aprendizaje automático aplican modelos de aprendizaje estadístico (gradient boosting, redes neuronales, Transformers) a series temporales financieras para predecir movimientos de precio o generar señales de trading. Estas implementaciones de código abierto cubren tanto formulaciones de regresión (objetivo de precio) como de clasificación (dirección).
13 algoritmos2 bibliotecas
Red de algoritmos
Cómo se conectan los algoritmos ML entre bibliotecas
ML Algorithms
Freqtrade9 algos
Microsoft Qlib6 algos
LightGBMRegressorintermediate
LightGBMClassifierintermediate
XGBoostRegressorintermediate
XGBoostClassifierintermediate
CatboostRegressorintermediate
PyTorchMLPRegressoradvanced
PyTorchTransformerRegressoradvanced
LGBModelintermediate
XGBModelintermediate
DNNModeladvanced
ALSTMadvanced
TFTModeladvanced
GATsadvanced
Pipeline de decisión de trading
Cómo cooperan los algoritmos ML en un sistema de trading
1
Market Data Input
Price, volume, indicators
OHLCV price data
Technical indicators
Feature engineering
2
ML Signal Generation
Model predictions
LightGBM/XGBoost regression
Transformer sequence modeling
MLP classification
3
Entry Decision
Signal threshold filtering
Confidence score > 0.6
Direction: Long/Short
4
Exit Strategy
Profit taking & stop loss
Take profit (fixed % or trailing)
Stop loss (ATR-based)
5
Risk Management
Position sizing & portfolio
Position sizing (Kelly criterion)
Portfolio rebalancing
Comparación de algoritmos
Comparar algoritmos ML en dimensiones clave
Matriz de comparación de algoritmosHaga clic en una columna para expandir los detalles
| Métrica | LightGBMRegressorFreqtrade | LightGBMClassifierFreqtrade | XGBoostRegressorFreqtrade | XGBoostClassifierFreqtrade | CatboostRegressorFreqtrade | PyTorchMLPRegressorFreqtrade | PyTorchTransformerRegressorFreqtrade | LGBModelQlib (Microsoft) | XGBModelQlib (Microsoft) | DNNModelQlib (Microsoft) | ALSTMQlib (Microsoft) | TFTModelQlib (Microsoft) | GATsQlib (Microsoft) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Complejidad | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced |
| Tipo de predicción | Regresión | Clasificación | Regresión | Clasificación | Regresión | Regresión | Regresión | Mixto | Mixto | Mixto | Secuencia | Mixto | Mixto |
| Velocidad de entrenamiento | ⚡⚡⚡ | ⚡⚡⚡ | ⚡⚡⚡ | ⚡⚡⚡ | ⚡⚡ | ⚡⚡ | ⚡ | ⚡⚡ | ⚡⚡ | ⚡⚡ | ⚡ | ⚡⚡ | ⚡⚡ |
| Precisión | 📊📊📊 | 📊📊📊 | 📊📊📊 | 📊📊📊 | 📊📊📊 | 📊📊📊 | 📊📊📊📊 | 📊📊 | 📊📊 | 📊📊 | 📊📊📊 | 📊📊 | 📊📊📊📊 |
| Ideal para | Datos tabulares | Datos tabulares | Datos tabulares | Datos tabulares | Uso general | Patrones no lineales | Patrones de series temporales | Uso general | Uso general | Uso general | Datos secuenciales | Uso general | Relaciones gráficas |
Freqtrade
Gradient boosting regression model for price movement prediction using LightGBM.
Velocidad⚡⚡⚡
Precisión📊📊📊
Parámetros clave
| n_estimators | 1000 | Number of boosting rounds |
| learning_rate | 0.01 | Step size shrinkage |
Origen:
freqai/prediction_models/LightGBMRegressor.pyGradient boosting classification model for directional prediction (up/down/neutral).
Velocidad⚡⚡⚡
Precisión📊📊📊
Parámetros clave
| n_estimators | 1000 | Number of boosting rounds |
Origen:
freqai/prediction_models/LightGBMClassifier.pyXGBoost-based regression model for continuous value prediction.
Velocidad⚡⚡⚡
Precisión📊📊📊
Parámetros clave
| n_estimators | 1000 | Number of boosting rounds |
| max_depth | 6 | Maximum tree depth |
Origen:
freqai/prediction_models/XGBoostRegressor.pyXGBoost-based classification model for directional prediction.
Velocidad⚡⚡⚡
Precisión📊📊📊
Parámetros clave
| n_estimators | 1000 | Number of boosting rounds |
Origen:
freqai/prediction_models/XGBoostClassifier.pyCatBoost gradient boosting model with native categorical feature support.
Velocidad⚡⚡
Precisión📊📊📊
Parámetros clave
| iterations | 1000 | Number of boosting iterations |
Origen:
freqai/prediction_models/CatboostRegressor.pyMulti-layer perceptron neural network for regression-based price prediction.
Velocidad⚡⚡
Precisión📊📊📊
Parámetros clave
| hidden_dim | 128 | Hidden layer dimension |
| dropout_percent | 0.2 | Dropout rate |
Origen:
freqai/prediction_models/PyTorchMLPRegressor.pyQlib (Microsoft)
LightGBM model for stock return prediction using technical and fundamental features.
Velocidad⚡⚡
Precisión📊📊📊
Parámetros clave
| num_leaves | 31 | Maximum number of leaves |
| learning_rate | 0.1 | Boosting learning rate |
Origen:
qlib/contrib/model/gbdt.pyXGBoost model for stock return prediction.
Velocidad⚡⚡
Precisión📊📊📊
Origen:
qlib/contrib/model/xgboost.pyDeep neural network for nonlinear feature extraction and return prediction.
Velocidad⚡⚡
Precisión📊📊📊
Parámetros clave
| hidden_size | 256 | Hidden layer size |
Origen:
qlib/contrib/model/pytorch_nn.pyAttention-based LSTM for sequential stock data modeling with attention mechanism.
Velocidad⚡
Precisión📊📊📊
Parámetros clave
| hidden_size | 64 | LSTM hidden size |
| num_layers | 2 | Number of LSTM layers |
Origen:
qlib/contrib/model/pytorch_alstm.pyTemporal Fusion Transformer combining static and temporal features for multi-horizon prediction.
Velocidad⚡⚡
Precisión📊📊📊
Origen:
qlib/contrib/model/pytorch_tft.pyAlgoritmos de trading con aprendizaje automático, referencia de algoritmos
- LightGBMRegressor (Freqtrade)
- Gradient boosting regression model for price movement prediction using LightGBM. Parámetros clave: n_estimators (Number of boosting rounds), learning_rate (Step size shrinkage).Origen: https://github.com/freqtrade/freqtrade/blob/develop/freqai/prediction_models/LightGBMRegressor.py.
- LightGBMClassifier (Freqtrade)
- Gradient boosting classification model for directional prediction (up/down/neutral). Parámetros clave: n_estimators (Number of boosting rounds).Origen: https://github.com/freqtrade/freqtrade/blob/develop/freqai/prediction_models/LightGBMClassifier.py.
- XGBoostRegressor (Freqtrade)
- XGBoost-based regression model for continuous value prediction. Parámetros clave: n_estimators (Number of boosting rounds), max_depth (Maximum tree depth).Origen: https://github.com/freqtrade/freqtrade/blob/develop/freqai/prediction_models/XGBoostRegressor.py.
- XGBoostClassifier (Freqtrade)
- XGBoost-based classification model for directional prediction. Parámetros clave: n_estimators (Number of boosting rounds).Origen: https://github.com/freqtrade/freqtrade/blob/develop/freqai/prediction_models/XGBoostClassifier.py.
- CatboostRegressor (Freqtrade)
- CatBoost gradient boosting model with native categorical feature support. Parámetros clave: iterations (Number of boosting iterations).Origen: https://github.com/freqtrade/freqtrade/blob/develop/freqai/prediction_models/CatboostRegressor.py.
- PyTorchMLPRegressor (Freqtrade)
- Multi-layer perceptron neural network for regression-based price prediction. Parámetros clave: hidden_dim (Hidden layer dimension), dropout_percent (Dropout rate).Origen: https://github.com/freqtrade/freqtrade/blob/develop/freqai/prediction_models/PyTorchMLPRegressor.py.
- PyTorchTransformerRegressor (Freqtrade)
- Transformer architecture for time-series regression using self-attention mechanism. Parámetros clave: num_heads (Number of attention heads), num_layers (Number of transformer layers).Origen: https://github.com/freqtrade/freqtrade/blob/develop/freqai/prediction_models/PyTorchTransformerRegressor.py.
- LGBModel (Qlib (Microsoft))
- LightGBM model for stock return prediction using technical and fundamental features. Parámetros clave: num_leaves (Maximum number of leaves), learning_rate (Boosting learning rate).Origen: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/gbdt.py.
- XGBModel (Qlib (Microsoft))
- XGBoost model for stock return prediction. Origen: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/xgboost.py.
- DNNModel (Qlib (Microsoft))
- Deep neural network for nonlinear feature extraction and return prediction. Parámetros clave: hidden_size (Hidden layer size).Origen: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/pytorch_nn.py.
- ALSTM (Qlib (Microsoft))
- Attention-based LSTM for sequential stock data modeling with attention mechanism. Parámetros clave: hidden_size (LSTM hidden size), num_layers (Number of LSTM layers).Origen: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/pytorch_alstm.py.
- TFTModel (Qlib (Microsoft))
- Temporal Fusion Transformer combining static and temporal features for multi-horizon prediction. Origen: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/pytorch_tft.py.
- GATs (Qlib (Microsoft))
- Graph Attention Networks modeling stock relationships for cross-stock prediction. Origen: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/pytorch_gats.py.