DNA codes over two noncommutative rings of order four

DNA codes based on error-correcting codes have been successful in DNA-basedcomputation and storage . Since there are four nucleobases in DNA, two wellknown algebraic structures such as the finite field $GF(4) and the integermodular ring $FF44$ have been used . In this paper, we describe a new type of DNA codes over two noncommutativerings $E$ and $F$ of order four with characteristic 2 .…

Strong Brascamp Lieb Inequalities

In this paper, we derive sharp nonlinear dimension-free Brascamp-Liebinequalities (including hypercontractivity inequalities) for distributions onPolish spaces, which strengthen the classic Brascamps-Liebs inequalities .Applications include the generalization of Mr. and Mrs. Gerber’s lemmas toR\’enyi entropies and distributions on Polish spaces .…

A Theoretical Performance Bound for Joint Beamformer Design of Wireless Fronthaul and Access Links in Downlink C RAN

It is known that data rates in standard cellular networks are limited due tointer-cell interference . An effective solution of this problem is to use themulti-cell cooperation idea . In Cloud Radio Access Network (C-RAN), which is acandidate solution in 5G and future communication networks, cooperation is applied by means of central processors (CPs) connected to simple remote radioheads with finite capacity fronthaul links .…

Sequential prediction under log loss with side information

The problem of online prediction with sequential side information underlogarithmic loss is studied . General upper and lower bounds on the minimaxregret incurred by the predictor is established . The upper bounds on theminimax regret are obtained by providing and analyzing a probability assignment inspired by mixture probability assignments in universal compression .…

Channel Estimation and Hybrid Combining for Wideband Terahertz Massive MIMO Systems

Terahertz (THz) communication is widely considered as a key enabler for 6G wireless systems . Massive multiple- input multiple-input multiple-output (MIMO) along with orthogonalfrequency division multiplexing (OFDM) can be used to deal with these problems . However, when the propagation delay across the base station (BS) antennaarray exceeds the symbol period, the spatial response of the BS array varies .…

Modelling Cooperation in Network Games with Spatio Temporal Complexity

The real world is awash with multi-agent problems that require collective action by self-interested agents . Such systems have local incentives for individuals whose behavior has an impact on the global outcome . Given appropriate mechanisms describing agent interaction, groups may achieve socially beneficial outcomes, even in the face of short-term self-interest incentives .…

A consistent and conservative Phase Field model for thermo gas liquid solid flows including liquid solid phase change

A consistent and conservative Phase-Field model is developed to study thermo-gas-liquid-solid flows with liquid-solid phasechange . The proposed model conserves the energy, preserves the temperature equilibrium, and isGalilean invariant . A novel continuous surface tension force to confine its contribution at the gas-liquid interface and a drag force modified from theCarman-Kozeny equation to reduce solid velocity to zero are proposed .…

Regularized Kaczmarz Algorithms for Tensor Recovery

Tensor recovery has recently arisen in a lot of application fields, such astransportation, medical imaging and remote sensing . Many tensor recovery methodshave been developed to apply various regularization techniques together with operator-splitting type of algorithms . In this work, we propose a novel algorithmic framework based on the Kaczmarzalgorithm for Tensor recovery .…

HAWKS Evolving Challenging Benchmark Sets for Cluster Analysis

Comprehensive benchmarking of clustering algorithms is rendered difficult by elusiveness of a unique mathematical definition of this unsupervised learning approach . We argue synthetic datasets must continue to play an important role in the evaluation of algorithms . We demonstratethe important role evolutionary algorithms play to support flexible generation of such benchmarks, allowing simple modification and extension.…

Self Reorganizing and Rejuvenating CNNs for Increasing Model Capacity Utilization

In this paper, we propose self-reorganizing and rejuvenating convolutionalneural networks . The proposed method utilizes thechannel activations of a convolution layer in order to reorganize that layersparameters . The reorganized parameters are clustered to avoid parameterredundancies . As such, redundant neurons with similar activations are merged leaving room for the remaining parameters to rejuvenate .…

LTL2Action Generalizing LTL Instructions for Multi Task RL

We address the problem of teaching a deep reinforcement learning (RL) agent to follow instructions in multi-task environments . We employ a well-known language — linear temporal logic (LTL) — to specify instructions, using a domain-specific vocabulary . We propose a novel approach to learning that exploits the compositional syntax and the semantics of LTL, enabling ourRL agent to learn task-conditioned policies that generalize to newinstructions, not observed during training .…

A Differentiable Contact Model to Extend Lagrangian and Hamiltonian Neural Networks for Modeling Hybrid Dynamics

The incorporation of appropriate inductive bias plays a critical role in learning dynamics from data . Real systems, such as legged robots and roboticmanipulators, involve contacts and collisions, which introduce discontinuities in the states . The proposed contact modelextends the scope of Lagrangian and Hamiltonian neural networks by allowingsimultaneous learning of contact properties and system properties .…

Deep Convolutional and Recurrent Networks for Polyphonic Instrument Classification from Monophonic Raw Audio Waveforms

Sound Event Detection and Audio Classification tasks are traditionally addressed through time-frequency representations of audio signals such asspectrograms . In this paper, we attempt to recognize musicalinstruments in polyphonic audio by only feeding their raw waveforms into deeplearning models . Various recurrent and convolutional architectures incorporating residual connections are examined and parameterized in order tobuild end-to-end classi-fiers with low computational cost and only minimalpreprocessing .…

Content Aware Speaker Embeddings for Speaker Diarisation

The content-aware speakerembeddings (CASE) approach is proposed . It extends the input of the speaker classifier to include not only acoustic features but also their correspondingspeech content, via phone, character, and word embeddings . Experimental results showed that CASE achieved a 17.8% relativespeaker error rate reduction over conventional methods .…

Asset Management in Machine Learning A Survey

Machine Learning (ML) techniques are becoming essential components of many systems today, causing an increasing need to adapt traditional engineering practices and tools to the development of ML-based systems . We discuss and position ML assetmanagement as an important discipline that provides methods and tools for ML assets as structures and the ML development activities as their operations .…

Why Security Defects Go Unnoticed during Code Reviews A Case Control Study of the Chromium OS Project

Peer code review has been found to be effective in identifying securityvulnerabilities . Despite practicing mandatory code reviews, many OpenSource software projects still encounter a large number of post-releasesecurity vulnerabilities . Alogistic regression model fitted on our dataset achieved an AUC score of 0.91and has identified nine code review attributes that influence identification of security defects .…

GenTree Using Decision Trees to Learn Interactions for Configurable Software

GenTree is a new dynamic analysis that automatically learns a program’s interactions – logical formulae that describe howconfiguration option settings map to code coverage . GenTree uses an iterativerefinement approach that runs the program under a small sample ofconfigurations to obtain coverage data; uses a custom classifying algorithm on these data to build decision trees representing interaction candidates; and then analyzes the trees to generate new configurations to further refine thetrees and interactions in the next iteration .…

ADEPT A Socio Technical Theory of Continuous Integration

ADEPT theory combines constructs that include humans, processes, documentation, automation and the project environment . Theory was derived from phenomena observed in previousempirical studies . It can be used to generate new propositions for future studies to understand continuous practices and their impact on the social and technical aspects of softwaredevelopment .…

Collisionless and Decentralized Formation Control for Strings

A decentralized feedback controller for multi-agent systems inspired by vehicle platooning is proposed . The closed-loop resulting from thedecentralized control action has three distinctive features: the generation of collision-free trajectories, flocking of the system towards a consensus state, and asymptotic convergence to a prescribed pattern of distances between agents .…

The Complexity of Transitively Orienting Temporal Graphs

In a temporal network with discrete time-labels on its edges, entities and information can only “flow” along sequences of edges whose time-labels arenon-decreasing (resp. increasing) along temporal paths . Anorientation of a temporal graph is called temporally transitive if, whenever $u$ has a directed edge towards $v $t_1$ and $v$ $w$ has adirected edge .…

Proof complexity of positive branching programs

Positive NBPs compute monotone Boolean functions, just like negation-free circuits orformulas, but constitute a positive version of (non-uniform) NL . The proof complexity of NBPs was investigated in previous work by Buss, Dasand Knop, using extension variables to represent the dag-structure .…

Barriers for recent methods in geodesic optimization

We study a class of optimization problems including matrix scaling, matrixbalancing, multidimensional array scaling, operator scaling, and tensor scaling . Trust region methods, which includebox-constrained Newton’s method, are known to produce high precision solutionsvery quickly for matrix scaling and matrix balancing .…

ReLU Neural Networks for Exact Maximum Flow Computation

Understanding the great empirical success of artificial neural networks is currently one of the hottest research topics in computer science . In this paper we study the expressive power of NNswith rectified linear units from a combinatorial optimization perspective . We show that, given a directed graph with $n$ nodes and $m$ arcs,there exists an NN of polynomial size that computes a maximum flow from any possible real-valued arc capacities as input .…

They Them Theirs Rewriting with Gender Neutral English

We perform a case study on thesingular they, a common way to promote gender inclusion in English . We define are-writing task, create an evaluation benchmark, and show how a model can betrained to produce gender-neutral English with <1% word error rate with nohuman-labeled data . We discuss the practical applications and ethical considerations of the task, providing direction for future work into inclusivenatural language systems. We discuss practical applications, ethical considerations, and provide guidance for future language systems . …

Supporting search engines with knowledge and context

We aim to support search engines in leveraging knowledge while accounting for context . We study how to make structured knowledge moreaccessible to the user when the search engine proactively provides such knowledge as context to enrich search results . We propose to model query resolution as a term classification task and propose a method to address it .…

Do as I mean not as I say Sequence Loss Training for Spoken Language Understanding

Spoken language understanding (SLU) systems extract transcriptions, as well as semantics of intent or named entities from speech . We propose non-differentiable sequence losses based on SLU metrics as a proxy for semantic error . We show that custom sequence loss training is the state-of-the-art on open SLU datasets and leads to 6% relative improvement in both ASR and NLU performance metrics on large proprietary datasets .…

Hybrid phonetic neural model for correction in speech recognition systems

Automatic speech recognition (ASR) often fail in environments that use language specific to particular domains . Some strategies have been explored to reduce errors inclosed ASRs through post-processing, particularly automatic spell checking, and deep learning approaches . In this article, we explore using a deep neuralnetwork to refine the results of a phonetic correction algorithm applied to atelesales audio database .…

Optimizing Inference Performance of Transformers on CPUs

The Transformer architecture revolutionized the field of natural languageprocessing (NLP) Transformers-based models (e.g., BERT) power many importantWeb services, such as search, translation, question-answering, etc. The optimizations are evaluated using theinference benchmark from HuggingFace, and are shown to achieve the speedup of up to x2.36 .…