Recommender Systems, Topological Data Analysis at Strata Conference

  • November 12, 2013 · 6:30 PM
  • This location is shown only to members

This is a FREE session hosted at Strata London Conference. Please note the address of venue and name of the room.

Talks start at 18.45

(Talk to be announced) Sr. Data Scientist @Pivotal

 

"Particles mining: Turning a mountain of data into a molehill" by Ellie Dobson, Application Engineer at Mathworks

The LHC experiment, based at CERN in Geneva, produces about a petabyte of data per second, but the front end of the Higgs discovery analysis was performed on a dataset that could fit on most laptops. In this talk I shall walk through strategies employed by particle physics experiments to search for complex patterns in an initial dataset too big to fit in any data warehouse. The first barrier of defense consists of a hardware-based trigger, backed up by a farm of software triggers running in real time. After the trigger reduces the initial input rate, the recorded data is replicated to a world wide computing cluster, where automatic software processing is performed to aggregate the binary data into fewer elements. The next stage is to run algorithms, trained on Monte Carlo simulation, to mine the data for promising looking signatures. Only after these steps are statistical analyses performed, which qualify whether the signatures seen can indeed be attributed to a new physics signal. Ellie spent most of her early life planning to be a musician but did a rather unexpected u-turn at the age of 18 and after a brief foray into teaching ended up reading physics at Oxford University. She was awarded her PhD in 2009 in particle physics, and spent many happy days hunting particles in the ATLAS detector as part of the LHC project. After embarking on a Marie Curie fellowship in affiliation with University College London, she recently took the plunge into the private sector and now works as an Application Engineer at MathWorks, specialising in parallel computing and data science.


"Topological Data Analysis: visual presentation of multidimensional data sets" by Edward Kibardin, Head of Data and Analytics @Base79.

Topology data analysis (TDA) is an unsupervised approach which may revolutionise the way data can be mined and eventually drive the new generation of analytical tools. The idea behind TDA is an attempt to "measure" shape of data and find compressed combinatorial representation of the shape. In ordinary topology, the combinatorial representations serve the purpose of providing the compressed representation of high dimensional data sets which retains information about the geometric relationships between data points. TDA can also can be used as a very powerful clustering technique. Edward will present the comparison between TDA and other dimension reduction algorithms like PCA, LLE, Isomap, MDS, and Spectral Embedding.

 

Break, Community Update and Data Science books giveaway.

 

Item Similarity Revisited by Mark Levy,  Sr. Data Scientist @Mendeley

The announcement of the Netflix Prize back in 2006 fired the starting pistol on a race to develop methods to predict preference based on collaborative filtering of ratings, a race which is still in progress, at least in academic circles. Netflix themselves commented on their tech blog in 2012 that predicted ratings form only a single, relatively uninfluential input feature to the model which they actually use to generate recommendations. Meanwhile several other industry players, particularly those whose datasets contain only implicit feedback and not ratings, are known still to use simple item similarity methods as the basis of their recommender systems.

Item similarity methods offer fast computation at recommendation time, and natural explanations of why particular items are being recommended, but they have not been a focus of academic research, except as benchmarks which can apparently be easily beaten by more complex algorithms, perhaps because item similarity tends to give high quality recommendations only when carefully tuned for a particular dataset. An interesting paper from 2012 bucked the trend by introducing Sparse Linear Methods (SLIM), and showing that they easily outperformed more complex preference prediction models for top-N recommendation, but at rather a high computational cost compared to traditional item similarity methods when applied to large datasets.

In this talk Mark will present experimental results which suggest that a simple relaxation of the problem constraints solved by SLIM can lead to an item similarity method which outperforms model-based algorithms but at reasonable computational cost. I put this in the context of some reflections on the reality of running large-scale industrial recommender systems based on experience at Last.fm and Mendeley, and also introduce a new open source Python software package implementing our version of SLIM and some other useful methods for working with implicit feedback data.


Special thanks to O’Reilly Strata Conference London for hosting our meetup and supporting our community.

Thanks to MongoDB, Cloudera, Pivotal for supporting our community



Join or login to comment.

  • Edward K.

    Thanks Carlos for organising this meetup, and everyone for such a nice reception of my talk about Topology Data Analysis.
    You can find my slides here: http://www.slideshare.net/infal...­

    I've also promised to release movie maps which I've generated from the Netflix data. Here you go:
    http://datarefiner.com/netflix1...­ - an image, where each point is a movie and links represent the movie similarities (3200x3200)
    http://datarefiner.com/netflix1...­ - an image where each movie point has a title annotation. Resolution: 17000x17000, Size: 86MB

    Please signup this website http://datarefiner.com­ if you wish to receive updates on TDA project developments and get a beta access when it will be released.

    9 · November 13, 2013

    • Edward K.

      Hi Paul, there is a very nice presentation about morse theory: http://www.math.fsu.e...­. If you want more, here the document written by one of the founders of TDA: http://www.ams.org/jo...­. If you have any more question please feel free to write me directly: [masked]

      1 · November 20, 2013

    • Paul

      Thanks Edward, this is really good. Thanks again for the presentation.

      1 · November 25, 2013

  • Mark B.

    Interesting event, and all the presentations were good. I don't think I'd ever expected to see the words 'muon', 'Higgs Boson' and 'Money Shot' all in the same presentation...but that's Data Science, I guess.

    3 · November 13, 2013

  • Kannappan S.

    The Large Hadron Collider talk was inspiring!

    2 · November 12, 2013

    • JUNIOR S.

      That talk was just amazing

      1 · November 13, 2013

  • JUNIOR S.

    This was a great meetup for sure, all the speakers were just excellent.
    Carlos you have done a great service to all

    1 · November 13, 2013

  • Milan

    Thanks for arranging Carlos. Slides for TDA and Recommender Systems will be appreciated.

    November 13, 2013

  • Dimitris P.

    Another great meetup!

    November 13, 2013

  • sahera k.

    It was the best meetup I have attend till now. Thank Carlo for arrang it. Thank for all speakers all of them were great. I think all of us need the slides.

    1 · November 13, 2013

  • felix o.

    Thanks Carlos I learned a lot

    November 13, 2013

  • Richard S.

    Thanks for arranging Carlos. A view of the slides would be appreciated.

    1 · November 13, 2013

  • John V.

    Thank you, Carlos, and a thank you to all the speakers for a night of great talks. I'd also like to see slides or notes made available, if possible. See you all next time.

    1 · November 13, 2013

    • Nick

      +1 would like slides

      2 · November 13, 2013

  • Antonio M.

    Amazing event! Thanks for the book!

    November 13, 2013

  • Heather S.

    Always learn something! Thanks to Carlos and sponsors. Would love to see slides.... can they go up?

    2 · November 13, 2013

  • Seref A.

    Thanks to everyone. Great talks, London underground made it a challenge to travel yesterday, but the effort was worth it. Looking forward to events Carlos mentioned.

    November 13, 2013

  • Cathy W.

    Thank you for a great meetup, lots to learn and just be inspired by. Compared to finding the Higgs Boson, my job seems easy!

    November 13, 2013

  • Charlie O.

    Great range of talks and topics. Thanks guys! (Maybe one talk too long - my concentration isn't that good after a long day…)

    November 12, 2013

  • Anahita

    Excellent meetup! Enjoyed the talks :)

    November 12, 2013

  • Nick

    excellent talks! very interesting and enjoyable mettup. enjoyed it very much, thanks for organising!

    November 12, 2013

  • TomH

    Four great talks - all of them very interesting. Well done to all the speakers, and to Carlos for organizing

    November 12, 2013

  • Nick T.

    Thanks Carlos - Really enjoyed tonight!

    November 12, 2013

  • A former member
    A former member

    Hi all,
    I developed a "data quality firewall" (in 5 years of hard work) and I'm looking for a business partner or interested company.

    1 · November 12, 2013

  • Sebastian C.

    Sorry cant make it but I am happy to see slides if any.

    November 12, 2013

  • Mark U.

    Thanks.

    November 12, 2013

  • Milan

    oreillyone

    2 · November 12, 2013

  • Roopesh

    Will not be able to make it. Facing issues in public transport.

    November 12, 2013

  • Mark U.

    Does anyone know the wifi password?

    November 12, 2013

  • yucel e.

    Sorry failed to make it as there have been massive problems with the underground. It's a great shame to miss this event!

    November 12, 2013

  • Andrey D.

    Problems with public transport :'( Will not be able to make it.

    November 12, 2013

  • A former member
    A former member

    Chaos in the Underground, no trains running. Won't gonna make it in time.

    November 12, 2013

  • Anwar H

    Sadly can't make it due to urgent last minute problem. Hope someone can still use my place.

    November 12, 2013

  • Herve

    Sadly can't make it, last minute issue w/ child care, quite frustrated as swimming on recommendation topic all day long and still want more! Updated status to release space if there is a waiting list.

    November 12, 2013

  • Federico Roman D.

    Yes!

    November 12, 2013

  • Ali S.

    See you all later today. Look forward to meetup and learn. cheers!

    November 12, 2013

  • sahera k.

    This will be very interesting presentation. Sure the meetup today will be crowded.

    November 12, 2013

  • Antonio M.

    Sorry, Can I come, today, I'm really interested on the talk about recommendation system and Netflix similarity algorithm! Please, some one drop me a line of confirm! Thanks!

    November 12, 2013

  • Carlos

    New talk added... How to collect&transform petabytes per second data and run algos trained on Monte Carlo simulations to datamine for interesting patterns at LHC- Large Hadron Collider, CERN. This is serious stuff done by Ellie!

    6 · November 11, 2013

  • Behailu

    I am looking forward for this conference, thanks

    November 11, 2013

  • Richard S.

    How does one find out where they are on the waitlist?

    3 · November 11, 2013

  • Jacopo

    It would be nice to leave commercial advertisement outside the meetup conversation.

    1 · November 7, 2013

    • Carlos

      The previous blatant spam message has been removed.

      2 · November 7, 2013

  • Pritesh P.

    Yes

    November 2, 2013

  • Suvendu b.

    Looking forward for this event

    November 1, 2013

  • sahera k.

    This is very interesting topic, Sure I will come.

    November 1, 2013

Our Sponsors

People in this
Meetup are also in:

Imagine having a community behind you

Get started Learn more
Rafaël

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