May 2018 - Dm-meeting - mailman.cs.vt.edu

Fwd: CATCS Newsletter (fwd)
by B. Aditya Prakash 25 May '18

25 May '18

FYI, many links seem useful! - Aditya --------- Forwarded message ---------- Date: Fri, 25 May 2018 09:47:35 -0700 (PDT) From: tim <tim(a)theory.stanford.edu> To: tim <tim(a)theory.stanford.edu> Subject: CATCS Newsletter Friends: you are receiving this newsletter because you kindly agreed to be the institutional contact for CATCS (https://thmatters.wordpress.com/). Please forward this message on to your theory-friendly colleagues. I expect to send at most 1-2 messages per year to this list. Here's an update on some of CATCS's recent activities, along with links to useful resources maintained at TheoryMatters: 1. David Woodruff compiled a list of major awards that are relevant to the theoretical computer science community. Included is a brief description of each, a link to the nomination instructions, and examples of researchers from TCS and related fields that won the award in the past. See https://thmatters.wordpress.com/tcs-awards-list/. 2. Shuchi Chawla is leading an effort to improve the quality of TCS entries on Wikipedia. She has put together an advice document for researchers who want to help and a spreadsheet where you can contribute suggestions for entries to be created or improved. See https://thmatters.wordpress.com/2017/05/02/tcs-wikipedia-project/. 3. Ryan O'Donnell collected advice on writing papers for general scientific venues like Science, Nature, and PNAS, from a number of TCS researchers who have successfully published in these venues in the past. See https://thmatters.files.wordpress.com/2017/05/nature-science-pnas-advice.pdf. 4. An oldie but a goodie: examples of successful TCS NSF Career proposals. See https://thmatters.wordpress.com/funding-opportunities-and-tips/career-examp… See https://thmatters.wordpress.com/ and the affiated sites https://cstheory-jobs.org/ and https://cstheory-events.org/ for still more resources. Don't forget that this year's TheoryFest is also the 50th anniversary of STOC! (Happening June 25-29, in Los Angeles.) See http://acm-stoc.org/stoc2018/ for details. Finally, I always welcome your suggestions for future CATCS projects. Tim

1 0

Thursday, May 24th, @ 12:00 pm, Torg 3160A, Anika Tabassum
by Sorour Ekhtiari Amiri 22 May '18

22 May '18

Please join us for our next DM-Meeting. ****DM-Meeting Spring 2018**** WHO: Anika Tabassum WHEN: Thursday, May 24th @ 12:00 pm WHERE: Torg 3160A WHAT: Anika will talk about the following paper, Title: Classifying multivariate time series by learning sequence-level discriminative patterns Abstract: Time series classification algorithms designed to use local context do not work on landcover classification problems where the instances of the two classes may often exhibit similar feature values due to the large natural variations in other land covers across the year and unrelated phenomena that they undergo. In this paper, we propose to learn discriminative patterns from the entire length of the time series and use them as predictive features to identify the class of interest. We propose a novel neural network algorithm to learn the key signature of the class of interest as a function of the feature values together with the discriminative pattern made from that signature through the entire time series in a joint framework. We demonstrate the utility of this technique on the landcover classification application of burned area mapping that is considerable societal importance. Best regards, Sorour

1 0

Thursday, May 17th, @ 12:00 pm, Torg 3160A, Anika Tabassum and Bijaya Adhikari
by Sorour Ekhtiari Amiri 15 May '18

15 May '18

Hi everyone, We rescheduled our today's DM-meeting for Thursday, May 17th. Please join us for our next DM-Meeting. ****DM-Meeting Spring 2018**** WHO: Anika Tabassum and Bijaya Adhikari *WHEN: Thursday, May 17th, @ 12:00 pmWHERE: Torg 3160A* WHAT: Anika and Bijaya will talk about the following papers, Title: MILE: A Multi-Level Framework for Scalable Graph Embedding Abstract: Recently there has been a surge of interest in designing graph embedding methods. Few, if any, can scale to a large-sized graph with millions of nodes due to both computational complexity and memory requirements. In this paper, we relax this limitation by introducing the MultI-Level Embedding (MILE) framework – a generic methodology allowing contemporary graph embedding methods to scale to large graphs. MILE repeatedly coarsens the graph into smaller ones using a hybrid matching technique to maintain the backbone structure of the graph. It then applies existing embedding methods on the coarsest graph and refines the embeddings to the original graph through a novel graph convolution neural network that it learns. The proposed MILE framework is agnostic to the underlying graph embedding techniques and can be applied to many existing graph embedding methods without modifying them. We employ our framework on several popular graph embedding techniques and conduct embedding for real-world graphs. Experimental results on five large-scale datasets demonstrate that MILE significantly boosts the speed (order of magnitude) of graph embedding while also often generating embeddings of better quality for the task of node classification. MILE can comfortably scale to a graph with 9 million nodes and 40 million edges, on which existing methods run out of memory or take too long to compute on a modern workstation. Title: Classifying multivariate time series by learning sequence-level discriminative patterns Abstract: Time series classification algorithms designed to use local context do not work on landcover classification problems where the instances of the two classes may often exhibit similar feature values due to the large natural variations in other land covers across the year and unrelated phenomena that they undergo. In this paper, we propose to learn discriminative patterns from the entire length of the time series and use them as predictive features to identify the class of interest. We propose a novel neural network algorithm to learn the key signature of the class of interest as a function of the feature values together with the discriminative pattern made from that signature through the entire time series in a joint framework. We demonstrate the utility of this technique on the landcover classification application of burned area mapping that is considerable societal importance. Best regards, Sorour

1 0

Wednesday, May 16th, @ 12:00 pm, Torg 3160A, Anika Tabassum and Bijaya Adhikari [LOCATION AND TIME CHANGED]
by Sorour Ekhtiari Amiri 14 May '18

14 May '18

Please join us for our next DM-Meeting. ****DM-Meeting Spring 2018**** WHO: Anika Tabassum and Bijaya Adhikari *WHEN: Wednesday, May 16th, @ 12:00 pmWHERE: Torg 3160A* WHAT: Anika and Bijaya will talk about the following papers, Title: MILE: A Multi-Level Framework for Scalable Graph Embedding Abstract: Recently there has been a surge of interest in designing graph embedding methods. Few, if any, can scale to a large-sized graph with millions of nodes due to both computational complexity and memory requirements. In this paper, we relax this limitation by introducing the MultI-Level Embedding (MILE) framework – a generic methodology allowing contemporary graph embedding methods to scale to large graphs. MILE repeatedly coarsens the graph into smaller ones using a hybrid matching technique to maintain the backbone structure of the graph. It then applies existing embedding methods on the coarsest graph and refines the embeddings to the original graph through a novel graph convolution neural network that it learns. The proposed MILE framework is agnostic to the underlying graph embedding techniques and can be applied to many existing graph embedding methods without modifying them. We employ our framework on several popular graph embedding techniques and conduct embedding for real-world graphs. Experimental results on five large-scale datasets demonstrate that MILE significantly boosts the speed (order of magnitude) of graph embedding while also often generating embeddings of better quality for the task of node classification. MILE can comfortably scale to a graph with 9 million nodes and 40 million edges, on which existing methods run out of memory or take too long to compute on a modern workstation. Title: Classifying multivariate time series by learning sequence-level discriminative patterns Abstract: Time series classification algorithms designed to use local context do not work on landcover classification problems where the instances of the two classes may often exhibit similar feature values due to the large natural variations in other land covers across the year and unrelated phenomena that they undergo. In this paper, we propose to learn discriminative patterns from the entire length of the time series and use them as predictive features to identify the class of interest. We propose a novel neural network algorithm to learn the key signature of the class of interest as a function of the feature values together with the discriminative pattern made from that signature through the entire time series in a joint framework. We demonstrate the utility of this technique on the landcover classification application of burned area mapping that is considerable societal importance. Best regards, Sorour

1 1

Wednesday, May 9th, @ 4:30 pm, McB 110, Anika Tabassum [LOCATION CHANGED]
by Sorour Ekhtiari Amiri 06 May '18

06 May '18

Please join us for our next DM-Meeting. ****DM-Meeting Spring 2018**** WHO: Anika Tabassum WHEN: Wednesday, May 9th, @ 4:30 pm *WHERE: McB 110*WHAT: Anika will talk about the following two papers, - 'Spatio-Temporal Neural Networks for Space-Time Series Forecasting and Relations Discovery' https://arxiv.org/pdf/1804.08562.pdf Abstract—We introduce a dynamical spatio-temporal model formalized as a recurrent neural network for forecasting time series of spatial processes, i.e. series of observations sharing temporal and spatial dependencies. The model learns these dependencies through a structured latent dynamical component, while a decoder predicts the observations from the latent representations. We consider several variants of this model, corresponding to a different prior hypothesis about the spatial relations between the series. The model is evaluated and compared to state-of-the-art baselines, on a variety of forecasting problems representative of different application areas: epidemiology, geo-spatial statistics, and car-traffic prediction. Besides these evaluations, we also describe experiments showing the ability of this approach to extract relevant spatial relations - Uncovering the Spatio-Temporal Dynamics of Memes in the Presence of Incomplete Information http://students.cse.tamu.edu/hge/papers/Ge_cikm16.pdf Abstract-Modeling, understanding, and predicting the spatio-temporal dynamics of online memes are important tasks, with ramifications on location-based services, social media search, targeted advertising and content delivery networks. However, the raw data revealing these dynamics are often incomplete and error-prone; for example, API limitations and data sampling policies can lead to an incomplete (and often biased) perspective on these dynamics. Hence, in this paper, we investigate new methods for uncovering the full (underlying) distribution through a novel spatio-temporal dynamics recovery framework which models the latent relationships among locations, memes, and times. By integrating these hidden relationships into a tensor-based recovery framework called AirCP - we find that high-quality models of meme spread can be built with access to only a fraction of the full data. Experimental results on both synthetic and real-world Twitter hashtag data demonstrate the promising performance of the proposed framework: an average improvement of over 27% in recovering the spatio-temporal dynamics of hashtags versus five state-of-the-art alternatives. Best regards, Sorour

1 0