Discussion of article "How to use ONNX models in MQL5" - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Reviews on applications (objects, models and tools)
1. survey on the application of deep learning in algorithmic trading (2021)
Algorithmic trading is one of the most concerned directions in financial applications.
Compared with traditional trading strategies, algorithmic trading applications perform forecasting and arbitrage with higher efficiency and more stable performance.
Numerous studies on algorithmic trading models using deep learning have been conducted to perform trading forecasting and analysis.
In this article, we firstly summarise several deep learning methods that have shown good performance in algorithmic trading applications, and briefly introduce some applications of deep learning in algorithmic trading.
We then try to provide the latest snapshot application for algorithmic trading based on deep learning technology, and show the different implementations of the developed algorithmic trading model.
Finally, some possible research issues are suggested in the future.
The prime objectives of this paper are to provide a comprehensive research progress of deep learning applications in algorithmic trading, and benefit for subsequent research of computer programme trading systems.
2. Deep Learning for Financial Applications : A Survey (2020)
Computational intelligence in finance has been a very popular topic for both academia and financial industry in the last few decades.
Numerous studies have been published resulting in various models. Meanwhile, within the Machine Learning (ML) field, Deep Learning (DL) started getting a lot of attention recently, mostly due to its outperformance over the classical models.
Lots of different implementations of DL exist today, and the broad interest is continuing. Finance is one particular area where DL models started getting traction, however, the playfield is wide open, a lot of research opportunities still exist.
In this paper, we tried to provide a state-of-the-art snapshot of the developed DL models for financial applications, as of today.
We not only categorised the works according to their intended subfield in finance but also analysed them based on their DL models.
In addition, we also aimed at identifying possible future implementations and highlighted the pathway for the ongoing research within the field.
3. Financial Time Series Forecasting with Deep Learning : A Systematic Literature Review: 2005-2019 (2019)
Financial time series forecasting is, without a doubt, the top choice of computational intelligence for finance researchers from both academia and financial industry due to its broad implementation areas and substantial impact.
Machine Learning (ML) researchers have come up with various models and a vast number of studies have been published accordingly.
As such, a significant amount of surveys exist covering ML for financial time series forecasting studies.
Lately, Deep Learning (DL) models started appearing within the field, with results that significantly outperform traditional ML counterparts.
Even though there is a growing interest in developing models for financial time series forecasting research, there is a lack of review papers that were solely focused on DL for finance.
Hence, our motivation in this paper is to provide a comprehensive literature review on DL studies for financial time series forecasting implementations.
We not only categorised the studies according to their intended forecasting implementation areas, such as index, forex, commodity forecasting, but also grouped them based on their DL model choices, such as Convolutional Neural Networks (CNNs), Deep Belief Networks (DBNs), Long-Short Term Memory (LSTM).
We also tried to envision the future for the field by highlighting the possible setbacks and opportunities, so the interested researchers can benefit
# Import necessary libraries import pandas as pd import numpy as np import matplotlib.pyplot as plt from sklearn.preprocessing import MinMaxScaler from tensorflow.keras.models import Sequential from tensorflow.keras.layers import Dense, LSTM, Dropout, Conv1D, MaxPooling1D, Flatten # Load the dataset df = pd.read_csv('https://query1.finance.yahoo.com/v7/finance/download/AAPL?period1=1411958400&period2=1636521600&interval=1d&events=history&includeAdjustedClose=true') # Preprocess the data scaler = MinMaxScaler(feature_range=(0, 1)) scaled_data = scaler.fit_transform(df['Close'].values.reshape(-1, 1)) # Split the data into training and testing sets training_size = int(len(scaled_data) * 0.7) testing_size = len(scaled_data) - training_size training_data = scaled_data[0:training_size, :] testing_data = scaled_data[training_size:len(scaled_data), :] # Function to create input features and labels for model def create_dataset(dataset, look_back): data_X, data_Y = [], [] for i in range(len(dataset)-look_back-1): a = dataset[i:(i+look_back), 0] data_X.append(a) data_Y.append(dataset[i + look_back, 0]) return np.array(data_X), np.array(data_Y) # Create input features and labels for training and testing sets look_back = 60 X_train, y_train = create_dataset(training_data, look_back) X_test, y_test = create_dataset(testing_data, look_back) # Reshape the input data for CNN-LSTM model X_train = np.reshape(X_train, (X_train.shape[0], X_train.shape[1], 1)) X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1], 1)) # Define and compile the CNN-LSTM model model = Sequential() model.add(Conv1D(filters=32, kernel_size=5, activation='relu', input_shape=(look_back, 1))) model.add(MaxPooling1D(pool_size=2)) model.add(Dropout(0.2)) model.add(Conv1D(filters=64, kernel_size=5, activation='relu')) model.add(MaxPooling1D(pool_size=2)) model.add(Dropout(0.2)) model.add(LSTM(50, return_sequences=True)) model.add(Dropout(0.2)) model.add(LSTM(50)) model.add(Dropout(0.2)) model.add(Dense(1)) model.compile(loss='mean_squared_error', optimizer='adam') # Train the model on the training data model.fit(X_train, y_train, epochs=100, batch_size=32) # Use the model to make predictions on the testing data y_pred = model.predict(X_test) # Rescale the data back to original values y_pred = scaler.inverse_transform(y_pred) y_test = scaler.inverse_transform(y_test.reshape(-1, 1)) # Plot the original and forecasted values plt.plot(y_test, label='Actual') plt.plot(y_pred, label='Predicted') plt.xlabel('Time') plt.ylabel('Close Price') plt.title('AAPL Stock Price Forecast') plt.legend() plt.show()Did you expect something different? Only through one's own experience and knowledge can there be a result.
I rather did not expect that such a thing is possible now, the technology is impressive. I agree that even with such tools, serious labour is required, and experience is the son of hard mistakes....
Transfer preprocessing to MT is not a problem, I really want to test the system in MT tester, in python/R I need to write my own tester, fraught with errors.
Using ONNX models in µl solves the implementation issue. However, not all models and not quite easy.
Training and optimisation of models is solved by a separate process in Python.
But of all the above, the first stage is the most time-consuming, highly creative and the most important. And it is impossible to realise it on µl. We don't consider primitive scaling as preprocessing. And the folk wisdom says: "Garbage in - rubbish out". There is too much to be additionally developed and implemented in MCL for full use of MO only on MCL. It is impossible to embrace the immense, especially since it is constantly expanding ...
Is it really that bad? After all, a lot of articles have already been written about neural networks. And there are even some where everything is "simple".....
But of all the above, the first stage is the most time-consuming, highly creative and the most important. And it is impossible to implement it on µl.
As far as I understand, we are talking about the difficulty of calculating some price functions that act as features of the network.
Can you give an example, perhaps there will be those here who will present an implementation of these things.
Is it really that bad? I mean, a lot of articles have already been written about neural networks. And there are even some where everything is "simple"....
It's not bad. Everything is in a very embryonic state compared to the modern development of preprocessing functionality.
I've been programming in R for quite a long time. I have been using Python not so long ago and the main thing I noticed while studying machine learning projects (packages) in Python is almost complete absence of preprocessing. The approach is that the model will take everything out. This is neither bad nor good. This is the approach that has developed historically.
In contrast, R has a huge arsenal of preprocessing with an understanding of what it's all for.
Although recently there has been a mutual borrowing of many useful developments.
I mean that it is impossible to cover the vastness, and developers strive to do it. To port what is written in C - yes, but to rewrite everything in MKL.....
Regarding ONNX. The area is new, very promising, rapidly developing and in demand. It means that there will be a possibility to drive into ONNX and preprocessing. TensorFlow already has layers of preprocessing, though simple, but still. You need to try it out, it will come in handy.
Of course it looks more attractive to use ONNX in training, but this is only for Linux. Maybe WSL experts can use it somehow.