Discussing the article: "Trend Prediction with LSTM for Trend-Following Strategies"
I don't get it: where is the regression2024.onnx model itself in the zip archive?
an_tar #:
I don't get it: where is the regression2024.onnx model itself in the zip archive?
I don't get it: where is the regression2024.onnx model itself in the zip archive?
Hello an_tar.
As mentioned in the article, this type of system is to be validated via rolling-window backtest. I didn't want to include all my trained model since 2008 to make the file heavy.
It is advised to use the framework introduced in the article to train your own model to be compatible with your personal validation method.
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
Check out the new article: Trend Prediction with LSTM for Trend-Following Strategies.
Long Short-Term Memory (LSTM) is a type of recurrent neural network (RNN) designed to model sequential data by effectively capturing long-term dependencies and addressing the vanishing gradient problem. In this article, we will explore how to utilize LSTM to predict future trends, enhancing the performance of trend-following strategies. The article will cover the introduction of key concepts and the motivation behind development, fetching data from MetaTrader 5, using that data to train the model in Python, integrating the machine learning model into MQL5, and reflecting on the results and future aspirations based on statistical backtesting.
Intuitively, trend-following strategies capitalize on gains in trending markets but perform poorly in choppy markets, where the strategy ends up buying at a premium and selling at a discount. Academic research has shown that classic trend-following strategies, such as the golden cross, work across multiple markets and timeframes over long periods of history. While these strategies may not be highly profitable, they have demonstrated consistent gains. Trend-following strategies typically profit from extreme outliers, which generate significantly higher profits than the average loss. The strategy’s tight stop-loss and "let profits run" approach result in a low win rate but a high reward-to-risk ratio per trade.
LSTM (Long Short-Term Memory) is a specialized type of recurrent neural network (RNN) designed to capture long-range dependencies in sequential data. It utilizes memory cells that can maintain information over long periods, overcoming the vanishing gradient problem that typically affects traditional RNNs. This ability to store and access information from earlier in the sequence makes LSTM particularly effective for tasks like time series forecasting and trend prediction. For regression problems, LSTM can model the temporal relationships between input features and predict continuous outputs with high accuracy, making it ideal for forecasting applications.
The motivation for this article is to leverage the power of LSTM for trend regression, predicting future trends and potentially filtering out bad trades that result from low trendiness. This motivation is based on the hypothesis that trend-following strategies perform better in trendy markets compared to detrended markets.
Author: Zhuo Kai Chen