@nbro ai.stackexchange.com/questions/4509/… this is not a duplicate. This is a good question which I would want to answer in a much better way, now that I am aware of few good resources. And this is the only question which touches topics not associated with any question of such type nor answered by anyone in the answer.

@DuttaA What defines a duplicate question is not just the post itself but the accepted answers too. That question and the linked one are basically asking the same thing "What do I need to study for machine learning" (one specifically asks for "mathematics", but, if you look at the accepted answers, they basically say the same thing, which means that the OPs were roughly looking for the same thing).

Even if that's not a duplicate, these questions are "too broad". Asking for 1. specific books (like this question) is ok (because people can focus on books), but asking "what to study for a field" is not only too broad (because people can give all types of answers, such as links to videos, books, lessons, articles, papers, etc.) but also a little bit primarily opinion-based.

You may argue that this is not an exact duplicate, but there's almost no exact duplicate, if you formulate your question differently.

However, let's analyze more closely why I closed them as duplicates. The duplicate question asks 3 apparently distinct questions (this is already a good reason to close it as too broad).

1. What do I need to study for machine learning? (title question, which is very vague)

2. Is it important to have my own expertise to work in the fields? (what does it mean "my own expertise"? which fields? I suppose that's just a typo; well, of course, it's important to have experience in any field to work in that field, so I don't understand this question)

3. If so, what kinds of subjects do I have to study right now? (Are they asking about specific sub-topics of ML or prerequisites? It's not clear to me, because the OP seems to be aware of many topics of ML, given that they list them in the post and they even say that they read the LSTM paper)

In the end, what's the question being asked here? When it's not clear exactly what's the question, I determine it from the accepted answer, which suggests "Pick an ML book". In other words, the question is "How do I get started in ML".

The other question has been edited by me to "save it" as a decent question. Let's only analyze the current version of the question, otherwise, I don't do anything more today, just by trying to explain why I closed the question.

Here there are also many questions

1. How can I start learning mathematics for machine learning?

2. Where should I start from?

3. Can someone provide me a road map for how to learn the mathematical background required for machine learning?

This question seems to focus on the "mathematical background", so one would expect answers like this or even yours (in the duplicate) to be accepted, but they accepted an answer that says "You should begin from Dr Andrew Ng machine learning course on Coursera. It's probably the most popular course for newcomers in machine learning. It's a free course.", i.e. the question was "How should I get started with ML".

So, the questions are really duplicates, to the extent that the accepted answers are very similar.

The problem with both questions was that they were both not specific enough and everyone started to give different types of answers (and that's often a good reason to close posts). From now on, when you see a similar problematic/vague question, you should ask for clarification "What are you looking for? Books, courses, articles, papers, video lessons, or what?" It's not ok to ask for all those things at the same time. Those questions are too broad.

In any case, if you don't agree with my decision, you can bring this topic to meta (of course, make sure to just focus on the issue, i.e. "post was closed but I don't think it should have been closed", and not me). In any case, I don't think these vague/broad questions are really beneficial, because maybe someone is looking for a book and he gets all types of answers.

Maybe someone else is looking for a course, and he gets a book, and so on and so forth. So, as a rule of thumb, if people ask this type of questions again, they should try to be more specific, such as "What are examples of free online courses that give an extensive overview of the ML field?". I don't think this question has already been asked, actually.