Multinomial Logistic Regression with Apache Spark

Logistic Regression can not only be used for modeling binary outcomes but also multinomial outcome with some extension. In this talk, DB will talk about basic idea of binary logistic regression step by step, and then extend to multinomial one. He will show how easy it's with Spark to parallelize this iterative algorithm by utilizing the in-memory RDD cache to scale horizontally (the numbers of training data.) However, there is mathematical limitation on scaling vertically (the numbers of training features) while many recent applications from document classification and computational linguistics are of this type. He will talk about how to address this problem by L-BFGS optimizer instead of Newton optimizer. 

Bio: DB Tsai is a machine learning engineer working at Alpine Data Labs. He is recently working with Spark MLlib team to add support of L-BFGS optimizer and multinomial logistic regression in the upstream. He also led the Apache Spark development at Alpine Data Labs. Before joining Alpine Data labs, he was working on large-scale optimization of optical quantum circuits at Stanford as a PhD student.




Join or login to comment.

  • Patrick N.

    Very useful presentation. The relative benefits of L-BFGS with Gauss-Newton was great. I also liked the use of aggregate as an alternative to map+reduce.

    June 21

  • al f.

    VIDEO: http://youtu.be/kiHrDEpsOEA

    (please wait for google to finish processing)

    3 · June 21

    • DB T.

      Thanks for recording!

      June 21

  • DB T.

    Thanks for coming on Friday night, and hope that you enjoy it. This is the slide I used tonight.
    http://www.slideshare.net/dbtsai/2014-0620-mlor-36132297

    Have a good night.

    4 · June 20

    • Shivani R.

      It was a wonderful talk. Great work and thanks for your patience in answering my questions

      1 · June 20

    • DB T.

      You are welcome. Nice question tho. Send me the log if you still have issue

      June 20

  • Mauricio R.

    Great content

    June 20

  • Emre

    Please who can't attend can at least study the slides: http://www.slideshare.net/dbtsai/2014-0501-mlor

    1 · June 20

  • Vlad S.

    Vote for video/stream, if it's possible.

    1 · June 20

  • Baswaraju

    Prefer to attend in person to keep this format live....

    June 20

  • Ram G.

    Yes please.

    June 20

  • Yu J.

    Votes up for video upload/live stream! 6:45pm on Friday night is not easy to make if not working in MTV. Thanks a lot!

    3 · June 20

  • Mahadev K.

    A video/ live stream would be great! Thanks :)

    4 · June 5

  • Raja K.

    I am interested and would like to participate if you have a live stream feed.

    2 · June 5

Our Sponsors

People in this
Meetup are also in:

Create a Meetup Group and meet new people

Get started Learn more
Allison

Meetup has allowed me to meet people I wouldn't have met naturally - they're totally different than me.

Allison, started Women's Adventure Travel

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy