Rolling Hashes, Compression Algorithms, and Mixture of Markov Chains

Update: Added a second talk

Another great event lined up for the next meetup. Same time, same place. See you there.

Please register on the Skillsmatter page as well.



Rolling Hashes and Compression AlgorithmsDr John Graham-Cumming (@jgrahamc)

Rolling hashes are fast to compute and can be used to identify 'similar' strings in blocks of data. This talk will will look at some uses of rolling hashes for string comparison and compression focussing on the rsync protocol, a spam filter, and a delta compression technique.

Bio: Dr John Graham-Cumming is an experienced programmer who works for CloudFlare, Inc. He is probably best known for having originated a successful petition to the British Government asking for an apology for its persecution of Alan Turing. He is author of The Geek Atlas, and started an organization whose aim is to build Charles Babbage's Analytical Engine.

Mixture of Markov Chains - Jose Llarena
Mixture of Markov Chains is a simple generative probabilistic model that can applied directly to raw sequential data like clickstreams or DNA sequences without any further pre-processing. In this talk I will describe its representation as a graphical model, how it relates to other probabilistic models, how to estimate its parameters via Expectation-Maximization, and how to use for clustering, data generation, density estimation and missing data completion. 

Bio: Jose Llarena has been writing intelligent software since 2001 in a variety of industries including fraud detection and e-learning.  He has obtained AI-related degrees from the Universities of Sussex and Edinburgh and is an active member of the London developer community.  

Join or login to comment.

  • Dirk G.

    As usual code/slides will appear in our github repo:

    November 11, 2013

  • Dirk G.

    Thanks all for coming and thanks Jose for a great talk. Lets hope John gets better soon and we will get him on for one of the next events.

    November 11, 2013

  • A former member
    A former member

    Would it be possible to organise events not on conflicting dates with the other major Big Data or data mining groups?

    November 11, 2013

    • Dirk G.

      As somebody who attended strata today I know what you mean :) I try to avoid conflicts but I am not always aware of what else is on. Also the date depends on my availability, the speaker's availability, and the venue's availability. The occasional clash is inevitable.

      November 11, 2013

  • Daniel P.

    Bleh, I really wanted to go. I just realized I have something else scheduled for today which I can't postpone. Have fun everyone!

    November 11, 2013

  • Dirk G.

    Quick reminder to ensure you have registered on the skillsmatter page if you are coming and to cancel your RSVP if you can no longer make it.

    November 10, 2013

  • Okash K.

    Will miss out due to software release date being pushed to Monday. Hopefully next time.

    November 9, 2013

  • Oliver C.

    Unfortunately I can't make this. Have fun everyone, hope someone can fill my spot :)

    November 8, 2013

  • Dirk G.

    Increased the number of spots for a second time. Now we are really up to capacity. Please remember to cancel if you can no longer make it.

    November 5, 2013

  • Andrzej G.

    Such a shame about the date, it's the devoxx week ;(

    November 4, 2013

  • Aran H.

    I'll be bringing myself and a fellow developer.

    October 18, 2013

  • Dirk G.

    Update: added a new talk on Mixture of Markov Chains by Jose Llarena for the 11/11 meetup. Make sure you register on the skills matter page as well.

    October 15, 2013

Our Sponsors

People in this
Meetup are also in:

Imagine having a community behind you

Get started Learn more

We just grab a coffee and speak French. Some people have been coming every week for months... it creates a kind of warmth to the group.

Rafaël, started French Conversation Group

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