Struc2vec: learning node representations from structural identity

September 15, 2017 ~ Adrian Colyer ~ 4 Comments

struc2vec: learning node representations from structural identity Ribeiro et al., KDD'17 This is a paper about identifying nodes in graphs that play a similar role based solely on the structure of the graph, for example computing the structural identity of individuals in social networks. That's nice and all that, but what I personally find most ... Continue Reading

Accelerating innovation through analogy mining

September 14, 2017 ~ Adrian Colyer ~ 1 Comment

Accelerating innovation through analogy mining Hope et al., KDD'17 Today's choice won a best paper award at KDD'17. It's a really interesting twist on information retrieval, building on a foundation of GloVe and word vectors to create purpose and mechanism vectors for a corpus of product descriptions. Using these vectors, the authors show how to ... Continue Reading

Adversarial examples for evaluating reading comprehension systems

September 13, 2017 ~ Adrian Colyer ~ 5 Comments

Adversarial examples for evaluating reading comprehension systems Jia & Liang, EMNLP 2017 We've now seen a number of papers investigating adversarial examples for images. In today's paper choice, Jia and Liang explore adversarial examples for text samples in the context of reading comprehension systems. The results are frankly a bit of a wake-up call for ... Continue Reading

Universal adversarial perturbations

September 12, 2017 ~ Adrian Colyer ~ 10 Comments

Universal adversarial perturbations Moosavi-Dezfooli et al., CVPR 2017. I'm fascinated by the existence of adversarial perturbations - imperceptible changes to the inputs to deep network classifiers that cause them to mis-predict labels. We took a good look at some of the research into adversarial images earlier this year, where we learned that all deep networks ... Continue Reading

Learning transferable architectures for scalable image recognition

September 11, 2017 ~ Adrian Colyer ~ Leave a comment

Learning transferable architectures for scalable image recognition Zoph et al., arXiv 2017 Things move fast in the world of deep learning! It was only a few months ago that we looked at 'Neural architecture search with reinforcement learning.' In that paper, Zoph et al., demonstrate that just like we once designed features by hand but ... Continue Reading

The ring of Gyges: investigating the future of criminal smart contracts

September 8, 2017 ~ Adrian Colyer ~ 5 Comments

The Ring of Gyges: investigating the future of criminal smart contracts Juels et al, CCS'16 The authors of this paper wrote it out of a concern for the potential abuse of smart contracts for criminal activities. And it does indeed demonstrate a number of ways smart contracts could facilitate crime. It's also though, another good ... Continue Reading

SmartPool: Practical decentralized mining

September 7, 2017 ~ Adrian Colyer ~ 1 Comment

SmartPool: Practical decentralized pooled mining Luu et al., USENIX Security 2017 Say you wanted to implement a mining pool that didn't place power in the hands of centralized pool operators. If only there was some fully decentralised way of establishing trust and coordinating activities according to a policy, in which anyone could participate... Oh, wait, ... Continue Reading

REM: Resource-efficient mining for blockchains

September 6, 2017 ~ Adrian Colyer ~ 7 Comments

REM: Resource-efficient mining for blockchains Zhang et al., USENIX Security 2017 The proof-of-work (PoW) used in most blockchains could just as easily be called proof-of-wasted-energy. All that hashing serves no useful end beyond electing the next block in the chain. The combined energy waste is actually pretty staggering: PoWs serve no useful purpose beyond consensus ... Continue Reading

Luck is hard to beat: the difficulty of sports prediction

September 5, 2017 ~ Adrian Colyer ~ Leave a comment

Luck is hard to beat: the difficulty of sports prediction Aoki et al., KDD'17 You can build all the outcome prediction models you like, such as the strategic play model we looked at yesterday, but some events have a certain amount of inherent unpredictability. There have been empirical and theoretical studies showing that unpredictability cannot ... Continue Reading

“The Leicester City fairytale?”: Utilizing new soccer analytics tools to compare performance in the 15/16 and 16/17 EPL seasons

September 4, 2017 ~ Adrian Colyer ~ 2 Comments

"The Leicester City Fairytale?" : Utilizing new soccer analytics tools to compare performance in the 15/16 and 16/17 EPL seasons Ruiz et al., KDD'17 In England the cricket season is coming to a close and a new football (soccer) season is getting underway. Today's paper choice is a bit of fun from the recent KDD'17 ... Continue Reading

the morning paper

a random walk through Computer Science research, by Adrian Colyer
Made delightfully fast by strattic

Month: September 2017