I don't think it's possible to learn "mathematical notation", as a thing. This is like asking "how can I learn human language".
There's different notation per field and subfield but I would argue that the process of figuring out notation per example or per paper is generalisable and not too difficult:
- identify the field of the paper; keywords, general categorisation, etc.
- take an example and break the notation into parts
- Google each of the parts independently alongside field keywords (results of this search should also give you some contextual info alongside each component which should help your knowledge)
- compile the aggregate of your research
- reread the paper and see if your understanding has improved
There's different notation per field and subfield but I would argue that the process of figuring out notation per example or per paper is generalisable and not too difficult:
- identify the field of the paper; keywords, general categorisation, etc.
- take an example and break the notation into parts
- Google each of the parts independently alongside field keywords (results of this search should also give you some contextual info alongside each component which should help your knowledge)
- compile the aggregate of your research
- reread the paper and see if your understanding has improved
- repeat
- try another paper in the same field