Everyone is welcome to ask questions, discuss and correct in the video bulletin and comments! Enter from here old version
Note video and note document series
Wu Enda is the best start!
- Deep Learning Series
- Wu Enda Deep Learning Course (5 lessons are all over): Open the smashed version
- Parsing job (no answer)
- CS231n Computer Vision Notes Summary
- CS294-112 depth enhanced learning (pre-2017fall version) Official website (2017fall version)
- CS224n NLP Official website bilibili
- NLP（oxford 2017） Official website
- Wu Enda Machine learning course
- Li Hongyi Official website address Machine learning course Deep learning course
Deep learning flower book notes (Read only when you are not reading)
I have been looking up to deep study papers and feel that way out of my reach, so I have been trying to avoid reading papers or even textbooks;
However, when I learned the CNN course, Wu Enda said that I can understand it, then try it, in case I really understand it…
Although the textbooks are very good, the authors who wrote the Deep Learning Book are all growing up…
In this content, Not only the notes for reading the paper, but also the notes for reading and interpreting the papers.
Read the source code of the deep learning library
Taking pytorch as an example, when we need to write a function that pytorch does not have, we need to provide forward and reverse transfer functions. The inverse function requires us to write the derivative calculation process ourselves; autograd can help us solve the problem of reverse homing. We only need to provide a forward function.
Learn the pytorch library
Learning keras library
Learning mxnet library
Learning scipy library
Learning seaborn library
Learn the Jupyter / IPython library
Exploring the method of video notes
- Some of the questions that are often asked Fragmentation problem
- Thinking and improving about video notes Update
- How to make video notes Initial attempt
joinquant uqer programming
What are the benefits of making your own tools?
Time-consuming, inefficient, full of mistakes, no use, no practical value
Self-education, figure out how each tool comes from, what are the characteristics, what are the specific small problems that can be solved; you can always debug to the bottom