How to master Machine Learning

22 June 2022, 11:25
MetaQuotes
0
2 797

All beginning traders start their learning journey with the technical analysis basics, and many of them read the same books on stock exchange trading. The basics are normally easy to understand. However, the initial manual trading phase passes fairly quickly. The next step is to achieve greater stability of trading results and to increase trading volumes, while covering a variety of financial instruments and maintaining low risk. This is where algorithmic trading via trading robots comes in handy, which is however a totally new area of study. In addition to financial market knowledge, it requires programming and technical analysis skills.

The algorithmic trading topic is too broad. By simply searching on the web, you can find hundreds or even thousands of specialized resources and tutorials. One of the approaches which is becoming increasingly popular concerns Machine Learning and Neural Networks. This is a whole new area of diverse knowledge, and thus it can be difficult to understand where to start learning this subject. We have prepared a selection of materials in English in order to save you time searching for this information.

Books


Books

A selection of books on how to use Machine Learning in algorithmic trading. This area requires knowledge of mathematics, statistics and Python programming skills.

  • Marcos López de Prado. Advances in Financial Machine Learning (Link)
  • Dr Howard B Bandy. Quantitative Technical Analysis: An integrated approach to trading system development and trading management (Link)
  • Tony Guida. Big Data and Machine Learning in Quantitative Investment (Link)
  • Michael Halls-Moore. Advanced Algorithmic Trading (Link)
  • Jannes Klaas. Machine Learning for Finance: Data algorithms for the markets and deep learning from the ground up for financial experts and economics (Link)
  • Stefan Jansen. Hands-On Machine Learning for Algorithmic Trading: Design and implement smart investment strategies to analyze market behavior using the Python ecosystem (Link)
  • Ali N. Akansu, Sanjeev R. Kulkarni , Dmitry M. Malioutov. Financial Signal Processing and Machine Learning (Link)
  • David Aronson. Evidence-Based Technical Analysis: Applying the Scientific Method and Statistical Inference to Trading (Link)
  • David Aronson. Statistically Sound Machine Learning for Algorithmic Trading of Financial Instruments (Link)
  • Ernest P. Chan. Machine Trading: Deploying Computer Algorithms to Conquer the Markets (Link)


Online Courses and Specializations

Online courses offer the most accessible and popular way to gain knowledge in focused areas. Here is a selection of Machine Learning courses available on Udacity and Coursera

Online Courses

  • Udacity: Georgia Tech. Machine Learning for Trading. This course introduces students to the real-world challenges of implementing machine learning based trading strategies. It focuses on how to apply probabilistic machine learning approaches to trading decision-making. The course covers statistical approaches like linear regression, KNN and regression trees and it considers how to apply them to actual stock trading situations. The duration is approx. 4 months.

  • Udacity: WorldQuant. Artificial Intelligence for Trading. This 6-month course provides the basics of quantitative analysis, including data processing, trading signal generation, and portfolio management. You will learn how to use Python to work with historical stock data, develop trading strategies, and construct a multi-factor model with optimization.

  • Coursera: NYU specialization Machine Learning and Reinforcement Learning in Finance Specialization (Weakly related to trading), which consists of four courses 13-24 hours each. This series is designed for three categories of students:

    1. Practitioners working at financial institutions such as banks, asset management firms or hedge funds.
    2. Individuals interested in applications of Machine Learning for personal day trading.
    3. Current full-time students pursuing a degree in Finance, Statistics, Computer Science, Mathematics, Physics, Engineering or other related disciplines who want to learn about practical applications of ML in Finance.
    Experience with Python (numpy, pandas, and IPython/Jupyter notebooks), linear algebra, basic probability theory and basic calculus is necessary to complete assignments in this course.

    The courses:
    • Guided Tour of Machine Learning in Finance. An introductory course with the focus on applications on Finance. 
    • Fundamentals of Machine Learning in Finance. A learner with some or no previous knowledge of Machine Learning will get to know main algorithms of Supervised and Unsupervised Learning, and Reinforcement Learning, and will be able to use ML open source Python packages to design, test, and implement ML algorithms in Finance.
    • Reinforcement Learning in Finance. The course aims at introducing the fundamental concepts of Reinforcement Learning (RL) and developing use cases for RL applications for option valuation, trading, and asset management.
    • Overview of Advanced Methods for Reinforcement Learning in Finance. The last course of the specialization provides a deeper look into topics discussed in the third course. In particular, it considers links between Reinforcement Learning, option pricing and physics, implications of Inverse Reinforcement Learning for modeling market impact and price dynamics, and perception-action cycles in Reinforcement Learning.


YouTube Videos

The list below shows some of the most useful trading videos concerning the application of Machine Learning.

Videos

  • Siraj Raval. Videos about stock market prediction using Deep Learning (Link). Cryptocurrency bot, Reinforcement Learning for stock prediction, TensorFlow and more.
  • QuantInsti YouTube. Webinars about Machine Learning for trading (Link). Machine Learning, Quantitative Finance, Optimal Portfolio Theory and many other educational videos.
  • Quantopian. Webinars about Machine Learning for trading (Link). The use of Classification and Regression in trading, Deep Learning, Big Data, creating models for sentiment analysis and many other videos on ML.
  • Sentdex. Machine Learning for Forex and Stock analysis and algorithmic trading (Link). An introductory course on machine learning and pattern recognition in forex and stock trading.
  • Sentdex. Python programming for Finance (videos include Machine Learning) (Link). Data preprocessing for machine learning using Python.
  • QuantNews. Machine Learning for Algorithmic Trading - three videos (Link). The videos describe the first steps in Machine Learning, Data Preparation and Hyper Parameter Tuning.
  • Howard Bandy. Machine Learning Trading System Development Webinar (Link). A recorded webinar on how to develop a trading system using Machine Learning.
  • Ernie Chan. Machine Learning for Quantitative Trading Webinar (Link). Another one hour webinar on Machine Learning.
  • Hitoshi Harada, CTO at Alpaca. Deep Learning in Finance Talk (Link). Deep Learning in finance from the CTO of the Alpaca company developing APIs to automate crypto and stock trading.
  • Prediction Machines. Deep Learning with Python in Finance Talk (Link). Speaker Ben Ball talks about how to implement a reinforcement learning algorithm in Python using TensorFlow, as well as presents background information around the deep learning algorithm class and the application to financial markets.
  • Master Thesis presentation, Uni of Essex. Analyzing the Limit Order Book, A Deep Learning Approach (Link). 
  • Tucker Balch. Applying Deep Reinforcement Learning to Trading (Link). Presentation of the final project in M.Sc. Algorithmic Trading at University of Essex. The goal of the project is to extract information from the Order Book Level 2 (DOM Level-2) using a Convolutional Neural Network.
  • Krish Naik. Stock Price Prediction And Forecasting Using Stacked LSTM-Deep Learning (Link). Constructing a machine learning model for stock market prediction. Stock market prediction is trying to determine the future value of a company stock or other financial instrument traded on a financial exchange.

    Blogs and Relevant Websites

    There are a lot of different Machine Learning related blogs and websites. Below are the most popular resources which might be useful for algorithmic trading purposes.

    Blogs

    • Quantstart. Machine Learning for Trading articles (Link). A few dozens of articles divided into categories. They cover theoretical foundations, mathematical formulas and examples of calculations in different programming languages.
    • Quantopian. Lecture notebooks on ML-related statistics (Link). The website is no longer available, but thanks to Ihor Marusyk you can read and watch all 56 lectures of this legendary resource.
    • AAA Quants. Tom Starke Blog (Link). The blog subject is "AI based solutions, quantitative analysis and data processing for finance." The well-structured series of articles, complimented with Python code, will give you an idea of how to apply mathematics in algorithmic trading.
    • RobotWealth. Kris Longmore Blog (Link). It is another useful Machine Learning blog. It also contains reviews of algorithmic trading books and articles. An extremely interesting resource.
    • Blackarbs blog (Link). An interesting blog by a trader who once crashed in discretionary trading and then switched to Python, quantitative analysis and fully automated trading. The blog features numerous publications since 2013, in which the trader shares his ideas and methods.
    • Hardikp, Hardik Patel blog (Link). Machine Learning for intraday trading, Neural Networks, Stock Prediction and more.


    Interviews

    Ten interviews on the application of machine learning in algorithmic trading. Leading industry experts and practitioners answer questions and share useful advice. The videos have automatically generated subtitles.

    Interviews

    • Chat with Traders EP042: Machine learning for algorithmic trading with Bert Mouler (Link)
    • Chat with Traders EP142: Algo trader using automation to bypass human flaws with Bert Mouler (Link)
    • Chat with Traders EP147: Detective work leading to viable trading strategies with Tom Starke (Link)
    • Chat with Traders Quantopian 5: Good Uses of Machine Learning in Finance with Max Margenot (Link)
    • Chat With Traders EP131: Trading strategies, powered by machine learning with Morgan Slade (Link)
    • Better System Trader EP023: Portfolio manager Michael Himmel talks AI and machine learning in trading (Link)
    • Better System Trader EP028: David Aronson shares research into indicators that identify Bull and Bear markets (Link)
    • Better System Trader EP082: Machine Learning With Kris Longmore (Link)
    • Better System Trader EP064: Cryptocurrencies and Machine Learning with Bert Mouler (Link)
    • Better System Trader EP090: This quants' approach to designing algo strategies with Michael Halls-Moore (Link)


    Scientific Papers

    Financial markets play an important role in the economic and social organization of modern society. Information is an invaluable asset in such markets. However, with the modernization of information systems, such a huge amount of data available to traders may make financial asset analysis difficult to impossible.

    Market researchers are developing intelligent methods and algorithms for decision support in various market segments. The list below contains more than 30 links to papers from scientific and educational institutions around the world. They cover Deep Learning, Classification and other AI topics in terms of their application to financial market prediction and trading.

    • Cumming, James. An Investigation into the Use of Reinforcement Learning Techniques within the Algorithmic Trading Domain (Link)
    • Marcos López de Prado. The 10 reasons most Machine Learning Funds fails (Link)
    • Xiong, Zhuoran et al. Practical Deep Reinforcement Learning Approach for Stock Trading (Link)
    • Ritter, Gordon. Machine Learning for Trading (Link)
    • Heaton, J.B. et al. Deep Learning for Finance: Deep Portfolios (Link)
    • Sirignano, Justin et al. Universal Features of Price Formation in Financial Markets: Perspectives From Deep Learning (Link)
    • Messmer, Marcial. Deep Learning and the Cross-Section of Expected Returns (Link)
    • Marcos López de Prado. Ten Financial Applications of Machine Learning (Presentation Slides) (Link)
    • Marcos López de Prado. The Myth and Reality of Financial Machine Learning (Presentation Slides) (Link)
    • Sepp, Artur. Machine Learning for Volatility Trading (Presentation Slides) (Link)
    • Marcos López de Prado. Market Microstructure in the Age of Machine Learning (Link)
    • Brogaard, Jonathan. Machine Learning and the Stock Market (Link)
    • Milan Fičura. Forecasting Foreign Exchange Rate Movements with k-Nearest-Neighbour, Ridge Regression and Feed-Forward Neural Networks (Link)
    • Edet, Samuel. Recurrent Neural Networks in Forecasting S&P 500 Index (Link). Hedayati, Amin et al. Stock Market Index Prediction Using Artificial Neural Network (Link)
    • Sen, Jaydip et al. A Robust Predictive Model for Stock Price Forecasting (Link)
    • Sezer, O.B. et al. An Artificial Neural Network-based Stock Trading System Using Technical Analysis and Big Data Framework (Link)
    • Singh, Ritika. Stock prediction using deep learning (Link)
    • Fischera Thomas, et al. Deep learning with long short-term memory networks for financial market predictions (Link)
    • Cavalcante,R.C. et al. Computational Intelligence and Financial Markets: A Survey and Future Directions (Link)
    • E. Chong et al. Deep Learning Networks for Stock Market Analysis and Prediction: Methodology, Data Representations, and Case Studies (Link)
    • Chien Yi Huang. Financial Trading as a Game: A Deep Reinforcement Learning Approach (Link)
    • W. Bao et al. A deep learning framework for financial time series using stacked autoencoders and longshort term memory (Link)
    • Zhou, Xingyu et al. Stock Market Prediction on High-Frequency Data Using Generative Adversarial Nets (Link)
    • Feng, Fuli et al. Improving Stock Movement Prediction with Adversarial Training (Link)
    • Z. Zhao et al. Time-Weighted LSTM Model with Redefined Labeling for Stock Trend Prediction (Link)
    • Arthur le Calvez, Dave Cliff. Deep Learning can Replicate Adaptive Traders in a Limit-Order-Book Financial Market (Link)
    • Dang Lien Minh et al. Deep Learning Approach for Short-Term Stock Trends Prediction Based on Two-Stream Gated Recurrent Unit Network (Link)
    • Yue Deng et al. Deep Direct Reinforcement Learning for Financial Signal Representation and Trading (Link)
    • Xiao Zhong. A comprehensive cluster and classification mining procedure for daily stock market return forecasting (Link)
    • J. Zhang et al. A novel data-driven stock price trend prediction system (Link)
    • Hoseinzade, Ehsan et al. CNNPred: CNN-based stock market prediction using several data sources (Link)
    • Chung, Hyejung et al. Genetic Algorithm-Optimized Long Short-Term Memory Network for Stock Market Prediction (Link)
    • Baek, Yujin et al. ModAugNet: A new forecasting framework for stock market index value with an overfitting prevention LSTM module and a prediction LSTM module ((Link)
    • Dash, Rajashree et al. A hybrid stock trading framework integrating technical analysis with machine learning techniques (Link)
    • E.A. Gerlein et al. Evaluating machine learning classification for financial trading: an empirical approach (Link)
    • Sirignano, Justin. Deep Learning for Limit Order Books (Link)


    News and Sentiment Trading

    There is a constant increase in the amount of news broadcast by various news agencies. The application of filters was required in order to benefit from this data stream therefore this feature was mainly used by the research departments of large investment firms. However, with the advent of news content digitization, developing computational capabilities and linguistic methods of interpretation, this data can now be analyzed efficiently and quickly. The programs that analyze this data are most commonly referred to as Sentiment Algorithms.

    • Frank Z. Xing et al. Natural language based financial forecasting: a survey (Link). The availability of data and various techniques leads to the qualitative development of Natural Language Processing (NLP) techniques. This increasing ability enables a more accurate capture of market sentiment.
    • Ziniu Hu et al. Listening to Chaotic Whispers: A Deep Learning Framework for News-oriented Stock Trend Prediction (Link). This article describes Hybrid Attention Networks, which are intended to predict stock trends based on a sequence of recent related events.
    • J.W. Leung, Master Thesis, MIT. Application of Machine Learning: Automated Trading Informed by Event Driven Data (Link). This article describes how to create trading strategies using
      machine learning methods for both technical analysis indicators and market sentiment. The described forecasting models can be used as algorithmic trading on any given stock exchange.
    • Xiao Ding et al. Deep Learning for Event-Driven Stock Prediction (Link). The article discusses Deep Learning methods for event-driven stock market prediction. First, events are extracted from the news text and are represented as dense vectors that are trained using a novel neural tensor network. Next, a deep convolutional neural network is used to model both short-term and long-term influences of events on stock price movements.


    Conclusion

    The purpose of this article is to provide traders with a brief, useful summary of publicly available machine learning tutorials. We hope that even a beginner will find something useful for themselves and get an insight of further development ideas. Some of the presented materials might require additional knowledge that goes far beyond a simple understanding of technical indicators and programming skills.

    If any of the subjects seem too complicated, now you know which courses to look for on the web in order to master the machine learning area. Learn something new, assist other traders, share links and your ideas in this complicated yet interesting area, via our MQL5.community.


    Translated from Russian by MetaQuotes Software Corp.
    Original article: https://www.mql5.com/ru/articles/10431

    DoEasy. Controls (Part 4): Panel control, Padding and Dock parameters DoEasy. Controls (Part 4): Panel control, Padding and Dock parameters
    In this article, I will implement handling Padding (internal indents/margin on all sides of an element) and Dock parameters (the way an object is located inside its container).
    MQL5 Wizard techniques you should know (Part 01): Regression Analysis MQL5 Wizard techniques you should know (Part 01): Regression Analysis
    Todays trader is a philomath who is almost always (either consciously or not...) looking up new ideas, trying them out, choosing to modify them or discard them; an exploratory process that should cost a fair amount of diligence. This clearly places a premium on the trader's time and the need to avoid mistakes. These series of articles will proposition that the MQL5 wizard should be a mainstay for traders. Why? Because not only does the trader save time by assembling his new ideas with the MQL5 wizard, and greatly reduce mistakes from duplicate coding; he is ultimately set-up to channel his energy on the few critical areas of his trading philosophy.
    Learn how to design a trading system by Ichimoku Learn how to design a trading system by Ichimoku
    Here is a new article in our series about how to design a trading system b the most popular indicators, we will talk about the Ichimoku indicator in detail and how to design a trading system by this indicator.
    Developing a trading Expert Advisor from scratch (Part 8): A conceptual leap Developing a trading Expert Advisor from scratch (Part 8): A conceptual leap
    What is the easiest way to implement new functionality? In this article, we will take one step back and then two steps forward.