ML offers algorithmic solutions and techniques that can be applied to many use cases. Parts 2, 3, and 4 of the book (as mentioned in Chapter 1, Machine Learning for Trading) have presented ML as a diverse set of tools that can add value to various steps of the strategy process, including:
- Idea generation and alpha factor research,
- Signal aggregation and portfolio optimization,
- Strategy testing
- Trade execution, and
- Strategy evaluation
Even more so, ML algorithms are designed to be further developed, adapted and combined to solve new problems in different contexts. For these reasons, it is important to understand key concepts and ideas underlying these algorithms, in addition to being able to apply them to data for productive experimentation and research as outlined in Chapter 6, The Machine Learning Process.
Furthermore, the best results are often achieved by combining human experts with ML tools. In Chapter 1, Machine Learning for Trading, we covered the quantamental investment style where discretionary and algorithmic trading converge. This approach will likely further grow in importance and depends on the flexible and creative application of the fundamental tools that we covered and their extensions to a variety of data sets.