Yes, definitely you can use machine learning to identify profitable puts to sell when trading options.
Machine learning is a field of Artificial Intelligence that provides systems which can find correlation by deep abstractions (and lots of computing power) to “learn” from experiences.
While there are quants across the world working on this, clearly no one has mastered such an algorithm. How do I know this? Well, if someone had a crystal ball (even with subpar accuracy) the markets could be fully exploited. There are rules preventing people from cornering the market and in effect it still stays relatively random.
Another principle of machine learning is the field of data sciences. This field works on finding correlation and patterns in very large data sets. It’s with these principles that many machine learning algorithms are based. E.g., if you had a machine learning program that was provided pictures of apples and oranges at various angle, lighting, etc. the machine would learn to find data that would attribute the picture to either an apple or an orange.
Another interesting field that has been around for quite some time is Multi-Criteria Decision Analysis (MCDA or MCDM) software. This allows algorithms to make decisions on subjective things. E.g., what is better an apple or an orange? Clearly there are multiple variables involved and are subjective to the person who will ultimately enjoy the particular piece of fruit. MCDM combined with principles from the data sciences and machine learning methods can yield very interesting results.
See instead of simply predicting up or down in a stock. These methods could make decisions on which particular option would best meet your outlook or particular trading style by comparing dissimilar datasets and understanding the preferences the trader has.
Apply your what you've learned with related Brutus Options Ranker components (Templates, Critieria, Market Groups, and Setups)