Search

Bilinear pooling has been recently proposed as a feature encoding layer,which can be used after the convolutional layers of a deep network, to improveperformance in multiple vision tasks. Instead of conventional global averagepooling or fully connected layer, bilinear pooling gathers 2nd...

The main observational equivalences of the untyped lambda-calculus have beencharacterized in terms of extensional equalities between B\"ohm trees. It iswell known that the lambda-theory H*, arising by taking as observables the headnormal forms, equates two lambda-terms whenever their B\"ohm trees...

We consider the problem of robustness in large consensus networks that occurin many areas such as distributed optimization. Robustness, in this context, isthe scaling of performance measures, e.g. H2-norm, as a function of networkdimension. We provide a formal framework to quantify the relation...

Software testing is difficult to automate, especially in programs which haveno oracle, or method of determining which output is correct. Metamorphictesting is a solution this problem. Metamorphic testing uses metamorphicrelations to define test cases and expected outputs. A large amount of time...

We address the problem of massive random access for an uncoordinated Gaussianmultiple access channel (MAC). The performance of $T$-fold irregular repetitionslotted ALOHA (IRSA) scheme for this channel is investigated. The maindifference of this scheme in comparison to IRSA is as follows: any...

We present Coconut, a novel selective disclosure credential scheme supportingdistributed threshold issuance, public and private attributes,re-randomization, and multiple unlinkable selective attribute revelations.Coconut can be used by modern blockchains to ensure confidentiality,authenticity and...

We propose a lightweight neural network model, Deformable Volume Network(Devon) for learning optical flow. Devon benefits from a multi-stage frameworkto iteratively refine its prediction. Each stage is by itself a neural networkwith an identical architecture. The optical flow between two stages...

Bioinformatics software plays a very important role in making criticaldecisions within many areas including medicine and health care. However, mostof the research is directed towards developing tools, and little time andeffort is spent on testing the software to assure its quality. In testing, atest...

Metamorphic testing is a well known approach to tackle the oracle problem insoftware testing. This technique requires the use of source test cases thatserve as seeds for the generation of follow-up test cases. Systematic design oftest cases is crucial for the test quality. Thus, source test case...

Different communities rely heavily on software, but use quite differentsoftware development practices. {\bf Objective}: We wanted to measure the stateof the practice in the area of statistical software for psychology tounderstand how it compares to best practices. {\bf Method}: We compared andranked...

We present an algebraic account of the Wasserstein distances $W_p$ oncomplete metric spaces. This is part of a program of a quantitative algebraictheory of effects in programming languages. In particular, we give axioms,parametric in $p$, for algebras over metric spaces equipped with...

Computing universal distributed representations of sentences is a fundamentaltask in natural language processing. We propose a method to learn suchrepresentations by encoding the suffixes of word sequences in a sentence andtraining on the Stanford Natural Language Inference (SNLI) dataset...

In many neural models, new features as polynomial functions of existing onesare used to augment representations. Using the natural language inference taskas an example, we investigate the use of scaled polynomials of degree 2 andabove as matching features. We find that scaling degree 2 features has...

We investigate the problem of detecting periodic trends within a string $S$of length $n$, arriving in the streaming model, containing at most $k$ wildcardcharacters, where $k=o(n)$. A wildcard character is a special character thatcan be assigned any other character. We say $S$ has wildcard-period $p...

Label propagation on the tensor product of multiple graphs can infermulti-relations among the entities across the graphs by learning labels in atensor. However, the tensor formulation is only empirically scalable up tothree graphs due to the exponential complexity of computing tensors. In thispaper...

We study secure and undetectable communication in a world where governmentscan read all encrypted communications of citizens. We consider a world wherethe only permitted communication method is via a government-mandated encryptionscheme, using government-mandated keys. Citizens caught trying to...

Coreset (or core-set) in this paper is a small weighted \emph{subset} $Q$ ofthe input set $P$ with respect to a given \emph{monotonic} function$f:\mathbb{R}\to\mathbb{R}$ that \emph{provably} approximates its fitting loss$\sum_{p\in P}f(p\cdot x)$ to \emph{any} given $x\in\mathbb{R}^d$. Using $Q$...

The paper describes a new algorithm to generate minimal, stable, and symboliccorrections to an input that will cause a neural network with ReLU neurons tochange its output. We argue that such a correction is a useful way to providefeedback to a user when the neural network produces an output that is...

We study a simple variant of the von Neumann model of an expanding economy,in which multiple producers produce goods according to their productionfunction. The players trade their goods at the market and then use the bundlesacquired as inputs for the production in the next round. We show that a...

We introduce a flexible class of well-quasi-orderings (WQOs) on words thatgeneralizes the ordering of (not necessarily contiguous) subwords. Each suchWQO induces a class of piecewise testable languages (PTLs) as Booleancombinations of upward closed sets. In this way, a range of regular...