About Me


E-mail: x@nyu.edu, where x=xy504

Sina Weibo: _____咖

GitHub: xingdi-eric-yuan

I’m currently a research engineer in Maluuba Research, I work in MRC team with some extremely smart guys.

I got my master’s degree in NYU (纽约大学), and bachelor’s degree in BJUT (北京工业大学)

I’m very interested in Machine Learning, Natural Language Processing, Computer Vision, and now I’m focusing on applying different kind of Deep Learning technics on helping machine to have better reading and reasoning skills.



Above are photos with Yann LeCun and  Rob Fergus. They are computer scientists I look up to with great respect.



  1. Posted May 6, 2014 at 1:43 pm | Permalink

    I enjoyed your blog and would like to talk to you regarding a project. Also probably about a summer internship doing an exciting project.

    Thank you
    Atul Hirpara

  2. Edwin
    Posted July 8, 2014 at 12:53 pm | Permalink


    I am a student from Denmark. I am currently trying to implement dropout for an autoencoder ? can I see the way you have tried (your code)? preferably in matlab.

  3. nikx
    Posted December 4, 2014 at 7:48 am | Permalink

    they are my inspiration too… seen a lot of videos of them… good clicks!!!

  4. 石文基
    Posted January 2, 2015 at 10:00 am | Permalink


    • Eric
      Posted January 2, 2015 at 3:03 pm | Permalink

      你好 🙂

  5. Posted January 14, 2015 at 10:53 pm | Permalink


  6. jsxyhelu
    Posted February 28, 2015 at 9:16 pm | Permalink

    Great job!

  7. achraf
    Posted July 13, 2015 at 6:22 am | Permalink


    It’s a very nice blog, I learned so much from it. I’m working on a project that do Handwriting recognition from bank chekcs, and I have problems locating the amount and the signature in the check with OpenCV, can you help me with this Please.

    Thanks in advance

  8. justin
    Posted October 8, 2015 at 3:27 am | Permalink

    Hi Dear;

    I am currently working on the implementation of an RBMs and have been facing some issues with the MNIST data reading. I actually use C++ in Unix Ubuntu but when trying to compile the sample codes for MNIST data you have posted online i get the following error:

    mnist2.cpp:14:1: error: invalid digit “8” in octal constant
    mnist2.cpp:15:33: fatal error: opencv2/core/core.hpp: No such file or directory
    compilation terminated.

    Wondering if you could help with some hints on how to solve the issue?
    Thanks and Kind Regards

  9. energy1010
    Posted October 9, 2015 at 2:45 am | Permalink

    I want to ask you a question about the cuda implementation of Rnn, how to use mini-batch

  10. mq
    Posted November 5, 2015 at 10:46 pm | Permalink

    It’s a nice blog, I can learned so much from it. I hope to have more opportunity to talk with you. I also come from China.

  11. Gudhi
    Posted March 16, 2016 at 10:57 pm | Permalink

    Hi Eric yuan

    I am Gudhi from USA, I really like your blog.
    I am working on a project , the Back propogation code which you posted will it work using MNIST Dataset. this is my doubt please clarify it.


    • Eric
      Posted March 17, 2016 at 9:59 am | Permalink

      Yes I think it works, and if you want to make sure, you can use some well-established frameworks like theano+keras, and it only uses less than 20 lines of code for doing the MNIST task. good luck!

  12. sasmita
    Posted June 21, 2016 at 6:00 am | Permalink

    i like ur blogs ,,thank you for sharing your knowledge

  13. amjad shah
    Posted August 12, 2016 at 2:59 am | Permalink

    hi sir
    i really like your blog.
    i am student of MS.i am working on CBIR. i need CNN Architecture for image classification for matlab.
    i search but did not understand.
    please help me it will be ur kindness..

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>