Algoritmos de trading com aprendizado de máquina
Modelos ML de código aberto para finanças quantitativas
Os algoritmos de trading com aprendizado de máquina aplicam modelos de aprendizado estatístico (gradient boosting, redes neurais, Transformers) a séries temporais financeiras para prever movimentos de preços ou gerar sinais de trading. Essas implementações de código aberto cobrem formulações de regressão (alvo de preço) e classificação (direção).
Como os algoritmos ML se conectam entre bibliotecas
Como os algoritmos ML cooperam num sistema de trading
Market Data Input
Price, volume, indicators
ML Signal Generation
Model predictions
Entry Decision
Signal threshold filtering
Exit Strategy
Profit taking & stop loss
Risk Management
Position sizing & portfolio
Comparar algoritmos ML em dimensões-chave
| Métrica | LightGBMRegressorFreqtrade | LightGBMClassifierFreqtrade | XGBoostRegressorFreqtrade | XGBoostClassifierFreqtrade | PyTorchMLPRegressorFreqtrade | PyTorchTransformerRegressorFreqtrade | LGBModelQlib (Microsoft) | XGBModelQlib (Microsoft) | DNNModelQlib (Microsoft) | ALSTMQlib (Microsoft) | TFTModelQlib (Microsoft) | GATsQlib (Microsoft) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Complexidade | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced | ⭐⭐⭐intermediate | ⭐⭐⭐intermediate | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced | ⭐⭐⭐⭐advanced |
| Tipo de previsão | Regressão | Classificação | Regressão | Classificação | Regressão | Regressão | Misto | Misto | Misto | Sequência | Misto | Misto |
| Velocidade de treino | ⚡⚡⚡ | ⚡⚡⚡ | ⚡⚡⚡ | ⚡⚡⚡ | ⚡⚡ | ⚡ | ⚡⚡ | ⚡⚡ | ⚡⚡ | ⚡ | ⚡⚡ | ⚡⚡ |
| Precisão | 📊📊📊 | 📊📊📊 | 📊📊📊 | 📊📊📊 | 📊📊📊 | 📊📊📊📊 | 📊📊 | 📊📊 | 📊📊 | 📊📊📊 | 📊📊 | 📊📊📊📊 |
| Ideal para | Dados tabulares | Dados tabulares | Dados tabulares | Dados tabulares | Padrões não lineares | Padrões de séries temporais | Uso geral | Uso geral | Uso geral | Dados sequenciais | Uso geral | Relações em grafo |
Freqtrade
Gradient boosting regression model for price movement prediction using LightGBM.
freqtrade/freqai/prediction_models/LightGBMRegressor.pyLightGBM classification model for user-defined discrete labels.
freqtrade/freqai/prediction_models/LightGBMClassifier.pyXGBoost-based regression model for continuous value prediction.
freqtrade/freqai/prediction_models/XGBoostRegressor.pyXGBoost classification model for user-defined discrete labels.
freqtrade/freqai/prediction_models/XGBoostClassifier.pyMulti-layer perceptron neural network for regression-based price prediction.
freqtrade/freqai/prediction_models/PyTorchMLPRegressor.pyQlib (Microsoft)
LightGBM model for stock return prediction using technical and fundamental features.
| num_leaves | 31 | Maximum number of leaves |
| learning_rate | 0.1 | Boosting learning rate |
qlib/contrib/model/gbdt.pyXGBoost model for stock return prediction.
qlib/contrib/model/xgboost.pyDeep neural network for nonlinear feature extraction and return prediction.
| hidden_size | 256 | Hidden layer size |
qlib/contrib/model/pytorch_nn.pyAttention-based LSTM for sequential stock data modeling with attention mechanism.
| hidden_size | 64 | LSTM hidden size |
| num_layers | 2 | Number of LSTM layers |
qlib/contrib/model/pytorch_alstm.pyTemporal Fusion Transformer combining static and temporal features for multi-horizon prediction.
qlib/contrib/model/pytorch_tft.pyAlgoritmos de trading com aprendizado de máquina, referência de algoritmos
- LightGBMRegressor (Freqtrade)
- Gradient boosting regression model for price movement prediction using LightGBM. Fonte: https://github.com/freqtrade/freqtrade/blob/c0a991f5632cb900075ff2864b6671ed71a319d7/freqtrade/freqai/prediction_models/LightGBMRegressor.py.
- LightGBMClassifier (Freqtrade)
- LightGBM classification model for user-defined discrete labels. Fonte: https://github.com/freqtrade/freqtrade/blob/c0a991f5632cb900075ff2864b6671ed71a319d7/freqtrade/freqai/prediction_models/LightGBMClassifier.py.
- XGBoostRegressor (Freqtrade)
- XGBoost-based regression model for continuous value prediction. Fonte: https://github.com/freqtrade/freqtrade/blob/c0a991f5632cb900075ff2864b6671ed71a319d7/freqtrade/freqai/prediction_models/XGBoostRegressor.py.
- XGBoostClassifier (Freqtrade)
- XGBoost classification model for user-defined discrete labels. Fonte: https://github.com/freqtrade/freqtrade/blob/c0a991f5632cb900075ff2864b6671ed71a319d7/freqtrade/freqai/prediction_models/XGBoostClassifier.py.
- PyTorchMLPRegressor (Freqtrade)
- Multi-layer perceptron neural network for regression-based price prediction. Fonte: https://github.com/freqtrade/freqtrade/blob/c0a991f5632cb900075ff2864b6671ed71a319d7/freqtrade/freqai/prediction_models/PyTorchMLPRegressor.py.
- PyTorchTransformerRegressor (Freqtrade)
- Transformer architecture for time-series regression using self-attention mechanism. Fonte: https://github.com/freqtrade/freqtrade/blob/c0a991f5632cb900075ff2864b6671ed71a319d7/freqtrade/freqai/prediction_models/PyTorchTransformerRegressor.py.
- LGBModel (Qlib (Microsoft))
- LightGBM model for stock return prediction using technical and fundamental features. Parâmetros principais: num_leaves (Maximum number of leaves), learning_rate (Boosting learning rate).Fonte: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/gbdt.py.
- XGBModel (Qlib (Microsoft))
- XGBoost model for stock return prediction. Fonte: 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. Parâmetros principais: hidden_size (Hidden layer size).Fonte: 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. Parâmetros principais: hidden_size (LSTM hidden size), num_layers (Number of LSTM layers).Fonte: 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. Fonte: 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. Fonte: https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/pytorch_gats.py.