Andrew Ng's advice for reading papers

很早之前自己记录的吴恩达阅读文章的建议，清电脑翻到的；相信各个领域读文章都差不多
找的一个完整版的中文翻译，比较全有其他内容
https://www.jiqizhixin.com/articles/2019-08-26-2

Guide principle

do not read every paper from the begin to the last word, try to quickly skim and figure out the content you read.

Specific steps:

1. title/abstract/figure (include caption, or maybe experiments)
2. intro. part (skim the related work in the first place)/conclusion/figure again
3. read the paper but skip the math
4. read the paper but skim the parts that do not make sense to you

Ask yourself the following questions after finishing reading:

1. what did the author try to accomplish ? (WHY the author decide to study this)
2. what the key elements of the approach ? (HOW the author solve the problem)
3. what can you use yourself ? (WHAT)
4. what other references you want to follow ?