https://blog.csdn.net/Linli522362242/article/details/102314389
Deep Neural Networks
Deep neural networks (DNNs) try to emulate the functioning of the human brain. They are in general composed of an input layer (the features), an output layer (the labels), and a number of hidden layers. The presence of hidden layers is what makes a neural network deep. It allows it to learn more complex relationships and to perform better on a number of problem types. When applying DNNs one generally speaks of deep learning instead of machine learning. For an introduction to this field, refer to Géron (2017) or Gibson and Patterson (2017).
DNNs with scikit-learn
This section applies the MLPClassifier algorithm from scikit-learn, as introduced in “Deep neural networks”. First, it is trained and tested on the whole data set, using the digitized features. The algorithm achieves exceptional performance in-sample (see Figure 15-17), which illustrates the power of DNNs for this type of problem. It also hints at strong overfitting, since the performance indeed seems unrealistically good:
from sklearn.neural_network import MLPClassifier
model = MLPClassifier(solver='lbfgs', alpha=1e-5, hidden_layer_sizes= 2*[250], random_state=1)
%time model.fit(data[cols_bin], data['direction'])
data['pos_dnn_sk'] = model.predict(data[cols_bin])
data['strat_dnn_sk'] = data['pos_dnn_sk'] * data['returns']
data[['returns', 'strat_dnn_sk']].sum().apply(np.exp)
data[['returns', 'strat_dnn_sk']].cumsum().apply(np.exp).plot(figsize=(10,6),
title='Performance of EUR/USD and DNN-based trading strategy (scikit-learn, in-sample)'
)
To avoid overfitting of the DNN model, a randomized train-test split is applied next. The algorithm again outperforms the passive benchmark investment and achieves a positive absolute performance (Figure 15-18). However, the results seem more realistic now:
DNNs with TensorFlow
TensorFlow has become a popular package for deep learning. It is developed and supported by Google Inc. and applied there to a great variety of machine learning problems. Zedah and Ramsundar (2018) cover TensorFlow for deep learning in depth
As with scikit-learn, the application of the DNNClassifier algorithm from TensorFlow to derive an algorithmic trading strategy is straightforward given the background from “Deep neural networks”. The training and test data is the same as before. First, the training of the model. In-sample, the algorithm outperforms the passive benchmark investment and shows a considerable absolute return (see Figure 15-19), again hinting at overfitting
train, test = train_test_split(data, test_size=0.5, random_state=100)
train = train.copy().sort_index()
test = test.copy().sort_index()
#Increases the number of hidden layers and hidden units.
model = MLPClassifier(solver='lbfgs', alpha=1e-5, max_iter=500, hidden_layer_sizes=3*[500], random_state=1)
%time model.fit(train[cols_bin], train['direction'])
test['pos_dnn_sk'] = model.predict(test[cols_bin])
test['strat_dnn_sk'] = test['pos_dnn_sk'] * test['returns']
test[['returns', 'strat_dnn_sk']].sum().apply(np.exp)
test[['returns', 'strat_dnn_sk']].cumsum().apply(np.exp).plot(figsize=(10,6),
title='Performance of EUR/USD and DNN-based trading strategy (scikit-learn, randomized train-test split)'
)
DNNs with TensorFlow
TensorFlow has become a popular package for deep learning. It is developed and supported by Google Inc. and applied there to a great variety of machine learning problems. Zedah and Ramsundar (2018) cover TensorFlow for deep learning in depth
As with scikit-learn, the application of the DNNClassifier algorithm from TensorFlow to derive an algorithmic trading strategy is straightforward given the background from “Deep neural networks”. The training and test data is the same as before. First, the training of the model. In-sample, the algorithm outperforms the passive benchmark investment and shows a considerable absolute return (see Figure 15-19), again hinting at overfitting
import tensorflow as tf
tf.logging.set_verbosity(tf.logging.ERROR)
fc =[tf.contrib.layers.real_valued_column('lags', dimension=lags)]
model = tf.contrib.learn.DNNClassifier(hidden_units=3*[500],
n_classes=len(bins)+1,
feature_columns=fc
)
def input_fn():
fc = {'lags': tf.constant(data[cols_bin].values)}
la = tf.constant(data['direction'].apply(lambda x: 0 if x<0 else 1).values,
shape=[data['direction'].size,1]
)
return fc, la
%time model.fit(input_fn=input_fn, steps=250)
model.evaluate(input_fn=input_fn, steps=1)
pred = np.array(list(model.predict(input_fn = input_fn)))
pred[:50]
data['pos_dnn_tf'] = np.where(pred>0,1,-1)
data['strat_dnn_tf'] = data['pos_dnn_tf'] * data['returns']
data[['returns', 'strat_dnn_tf']].sum().apply(np.exp)
data[['returns', 'strat_dnn_tf']].cumsum().apply(np.exp).plot(figsize=(10,6),
title='Performance of EUR/USD and DNN-based trading strategy (TensorFlow, in-sample)'
)
The following code again implements a randomized train-test split to get a more realistic view of the performance of the DNN-based algorithmic trading strategy. The performance is, as expected, worse out-of-sample (see Figure 15-20). In addition, given the specific parameterization the TensorFlow DNNClassifier underperforms the scikit-learn MLPClassifier algorithm by quite few percentage points:
model = tf.contrib.learn.DNNClassifier(hidden_units=3*[500],
n_classes=len(bins)+1,
feature_columns=fc
)
data=train
%time model.fit(input_fn=input_fn, steps=2500)
data=test
model.evaluateate(input_fn=input_fn, steps=1)
pred = np.array(list(model.predict(input_fn=input_fn)))
test['pos_dnn_tf'] = np.where(pred>0,1,-1)
test['strat_dnn_tf'] = test['pos_dnn_tf']*test['returns']
test[['returns','strat_dnn_sk', 'strat_dnn_tf']].sum().apply(np.exp)
test[['returns','strat_dnn_sk','strat_dnn_tf']].cumsum().apply(np.exp).plot(figsize=(10,6),
titile='Performance of EUR/USD and DNN-based trading strategy (TensorFlow, randomized train-test split)')