addressalign-toparrow-leftarrow-rightbackbellblockcalendarcameraccwcheckchevron-downchevron-leftchevron-rightchevron-small-downchevron-small-leftchevron-small-rightchevron-small-upchevron-upcircle-with-checkcircle-with-crosscircle-with-pluscontroller-playcrossdots-three-verticaleditemptyheartexporteye-with-lineeyefacebookfolderfullheartglobegmailgooglegroupshelp-with-circleimageimagesinstagramFill 1light-bulblinklocation-pinm-swarmSearchmailmessagesminusmoremuplabelShape 3 + Rectangle 1ShapeoutlookpersonJoin Group on CardStartprice-ribbonprintShapeShapeShapeShapeImported LayersImported LayersImported Layersshieldstartickettrashtriangle-downtriangle-uptwitteruserwarningyahoo

Mr. Stock Market, meet Mr. Machine Learning

In this Meetup, Dan Bikle presents  the use of machine learning for stock market predictions.  He will explore the following topics:

  - Read Stock Market Data off the web
  - Read Stock Market Data from the Interactive Brokers API
  - Load the data into a transformation environment
  - Extract interesting Machine Learning features:
     - Normalized Gains
     - Lagging Deltas
     - Moving Averages
     - RSI Values
     - Morning Gaps
     - Day Gaps
     - Daily close to close gap
     - Weekly close to close gap
     - Monthly close to close gap
  - Explore the features for interesting heuristics
  - This a feature or y-value? Ah-ha, it is both!
  - Feed the features to Machine Learning Algorithms
  - Back testing methodologies
  - Testing predictions in stock market tournaments

He will also discuss questions such as “If my algorithm succeeds on past data, will it work in the future?” and “If my algorithm fails on past data, will it fail in the future?” 

This Meetup will focus on predictions and heuristic based probabilities.

Discussions about transforming the predictions into optimal trading strategies is way beyond the scope of this Meetup.

Lastly, Dan will hold a free hands-on workshop at a later date outside of the meetup for those interested in digging into stock market data.

Disclosure:  Neither Dan nor the meetup organizers can assure that you will make money (or not lose money) through the techniques and strategies discussed during the meetup.  However, it will be interesting to learn. 

Speaker Bio: 

Dan is an Independent Consultant with a focus on Software Development.
He holds a BSEE from Caltech.
Contact Dan: [masked]





Join or login to comment.

  • Dan B.

    Class content: Tips for installing MADlib: http://www.spy611.com/blog/class

    May 4, 2014

  • Dan B.

    Tips for installing MADlib: http://www.spy611.com/blog/madlib

    May 4, 2014

  • Robert S.

    Great Meet up!
    Dan has spent years studying the stock market combined with machine learning algorithms.
    I am porting Dans awesome work to R and am looking for others who wish to contribute during his next classes.
    http://www.meetup.com/Palo-Alto-Data-Science-Association/events/178749252/
    His class uses Centos and Postgres and the associated MADLib statistical library. I think R may be an simpler choice. I created the first feature set in R during the last class and will work on adding his machine learning algorithms during this and next weeks class. Here is the R I wrote last week: https://github.com/danbikle/spy6...­
    If you know some R and wish to apply it to stocks join this Saturday and help create the R port.

    1 · April 28, 2014

    • Dan B.

      The next PADS Meetup will attract a combination of people who know R, Big-Data, and ML. Hopefully we can attract some of them to our Stock Market efforts: http://www.meetup.com...­

      1 · April 28, 2014

  • Patrick N.

    Would you be interested in looking at income statement, balance sheet & cash flow to screen companies to buy & hold?

    1 · April 28, 2014

    • Dan B.

      I like this idea. We should figure out how to get access to this data as it existed in the past so we can learn from it. Any data which can be connected to elements in a time-series should be studied for usefulness.

      2 · April 28, 2014

  • Dan B.

    Awesome. Dan is both knowledgeable and detailed. No powerpoint, just code, concepts and questions for the crowd.

    1 · April 27, 2014

  • Dan B.

    Crowd,

    we have 120 RSVPs and 228 on the waitlist.

    If you have RSVP but are not going please cancel.

    If you are attending, arrive early.

    The room holds 98; LATECOMERS GET NO SEAT.

    If you arrive too late for a seat,
    just relax at Hacker Dojo and use the free WIFI.

    Maybe you could get an early start on the homework
    for Dan's Machine Learning class:

    http://www.meetup.com/Palo-Alto-Data-Science-Association/events

    Also the Hacker Dojo membership considers it polite if you give an e-mail address to the computer near the entrance.

    Also they consider if polite if you toss $1 in the donation box near the entrance.

    The $1 pays for rent and WIFI.

    About parking, we have LOTS of free parking next door at Peets.

    Thanks,

    -- Dan

    1 · April 25, 2014

  • Ramya N.

    Unfortunately, I will not be able to make it to the session today. Changing my RSVP to No, hope it helps someone else equally interested.

    April 25, 2014

  • Tom M.

    Will there be any live streaming of the event?

    1 · April 17, 2014

  • veena

    Thanks Dan. I will check it out in the afternoon. Do I need API to connect to IB ?

    1 · April 24, 2014

    • Dan B.

      IB (Interactive Brokers) API is a big topic. The main obstacle is getting an account with them. Account costs money. After you get the account, you need to go through the process of upgrading your privileges. At the very least, it is a 1 hour project.

      1 · April 24, 2014

  • veena

    Can I do the coding on a mac?

    1 · April 23, 2014

    • Dan B.

      On mac anything is possible. CentOS is better than mac for 2 reasons: It is easier. It is the OS you would use for 'production'. Here is info about installing MADlib on all types of OS: https://github.com/mad...­

      1 · April 23, 2014

    • Steve

      All that said it is not uncommon for trading firms (especially non-banks) to develop on OSX and deploy on either linux or freebsd.

      1 · April 24, 2014

  • Dan B.

    The class associated with this material has been announced. I invite you to the class which is free and open to the public:
    http://www.meetup.com/Palo-Alto-Data-Science-Association/events/178853232/

    1 · April 23, 2014

  • Dan B.

    I have finished the slides. You will find them here: http://www.spy611.com/blog/preso

    2 · April 22, 2014

  • Tony T.

    In case anyone is interested, the SF BayArea ML group is also putting together finance+ML workshop soon (5/8/2014): http://www.meetup.com/SF-Bayarea-Machine-Learning/events/177809302/

    1 · April 21, 2014

  • Mohsen S.

    I am a Data Scientist interested in learning more about application of ML in finance

    1 · April 19, 2014

  • Dan B.

    This Meetup is now full. If you are not attending, PLEASE cancel your RSVP to make room for the reliable people.

    1 · April 17, 2014

  • Dan B.

    If you have a nice car,
    avoid Hacker Dojo parking lot and street out front.
    Instead park next door at Peets on Ellis.
    They close at 6pm and have lots of parking.
    Take 2 spots if you want.
    We have permission to park there.

    2 · April 17, 2014

  • Rahul

    Dan, will you be presenting using scikit-learn and madlib (as described in your blog) for this meetup?

    1 · April 17, 2014

    • Dan B.

      Yes I will discuss the technology. Also if you want to wrestle with the tech, I will announce a hands-on event at a later date. Have your laptops ready!

      1 · April 17, 2014

  • Dan B.

    Pre Lecture Reading Assignment: spy611.com/blog

    1 · April 17, 2014

People in this
Meetup are also in:

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