Transfer Learning for Human & AI

Ripple library provides a unified view of the computational space across multiple dimensions and multiple GPUs . It allows polymorphic datalayout, and provides a simple graph interface to describe an algorithm fromwhich inter-GPU data transfers can be optimally scheduled . We describe the abstractions provided by Ripple to allow complex computations to be describedsimply, and to execute efficiently across many GPUs with minimal overhead .…

Double sided auctions are widely used in financial markets to match demand and supply . We extend notions of double sided auctions to incorporate multiple quantity trade requests . We establish new uniqueness theorems that enable automatic detection ofviolations in an exchange program by comparing its output with that of averified program .…

To maintain blockchain-based services with ensuring its security, it is animportant issue how to decide a mining reward so that the number of minersparticipating in the mining increases . We propose a dynamical model ofdecision-making for miners using an evolutionary game approach .…

There is a trivial algorithm for approximatetriangle counting where $T$ is the number of triangles in the graph and $n$ thenumber of vertices . At the same time, one may count triangles exactly usingfast matrix multiplication in time $O . We answer this question positively by providing an algorithm which runs in time .…

Paper mainly focuses on the Ponzi scheme, atypical fraud, which has caused large property damage to the users in Ethereum . It proposes a detecting model based on graph convolutionalnetwork (GCN) to precisely distinguishPonzi contracts . Experiments on differentreal-world . datasets demonstrate that our proposed model has promising results compared with .…

In recent years, conversational agents have provided a natural and convenient access to useful information in people’s daily life . The main challenge is how to well capture and fullyexplore the historical context in conversation to facilitate effectivelarge-scale retrieval . We propose a novel graph-guided retrieval method to model the relations amonganswers across conversation turns .…

LiDAR depth completion is a task that predicts depth values for every pixel on the corresponding camera frame . Most of the existing state-of-the-art solutions are based on deepneural networks . In this letter, a novel non-learning depth completionmethod is proposed by exploiting the local surface geometry that is enhanced by an outlier removal algorithm .…

Fraudulent activity associated with public procurement contracts accounts for losses of billions of eurosevery year . Analysts use network science methods to study the co-biding relationships between firms that participate in publictenders issued by the $184$ municipalities of the State of Cear\’a (Brazil) between 2015 and 2019 .…

Answer selection (AS) is an essential subtask in the field of naturallanguage processing with an objective to identify the most likely answer to agiven question from a corpus containing candidate answer sentences . ASBERT is a framework built on the BERT architecture that uses Siamese and Triplet neural networks to learn an encoding function that maps a text to a fixed-size vector in an embedded space .…

BERT-based text ranking models have dramatically advanced the state-of-the-art in ad-hoc retrieval . Co-BERT has been proposed to exploit several BERT architectures to calibrate the query-documentrepresentations using pseudo relevance feedback before modeling the relevance of a group of documents jointly .…

We propose a periodic channel hopping sequence called PPoL (Packing thePencil of Lines in a finite projective plane) for the multichannel rendezvousproblem . We construct CH sequences with the maximumtime-to-rendezvous (MTTR) bounded by $N^2+3N+3$ If the number of commonly available channels is at least two .…

Deep Click-Through Rate (CTR) models are deployed in the commercial systems by industrial companies . To achieve better performance, it is necessary to train the deep CTR models on huge volume of training data efficiently . The training data for CTR models is usuallyhigh-dimensional and sparse .…

High-Efficiency Video Coding (HEVC) surpasses its predecessors in encoding efficiency by introducing new coding tools at the cost of an increased encodingtime-complexity . The Coding Tree Unit (CTU) is the main building block used in HEVC . In the HEVC standard, frames are divided into CTUs with the predeterminedsize of up to 64×64 pixels .…

Document layout comprises both structural and visual (eg. font-sizes)information that is vital but often ignored by machine learning models . We propose a novel layout-aware multimodalhierarchical framework, LAMPreT, to model the blocks and the whole document . We evaluate the proposed model on twolayout-aware tasks — text block filling and image suggestion — and show the effectiveness of our proposed hierarchical architecture as well as pretrainingtechniques .…

The Competition on Legal Information Extraction/Entailment (COLIEE-2019) run in association with the International Conference on Artificial Intelligence andLaw (ICAIL)-2019 has come up with few challenging tasks . We obtain encouraging results in all these four sub-tasks (i.e. Task1, Task2, Task3 and Task4), which will be able toprovide few automated systems to the judicial system .…

We consider the multi-broadcast problem in arbitrary connected radio networks . Each source node has a distinct piece of information that it wants to share with all nodes in the network . We set out todetermine the shortest possible labels so that the problem can be solveddeterministically in the labeled radio network by some universal deterministic algorithm .…

This paper introduces a high-quality open-source speech synthesis dataset for Kazakh . The dataset consists of about 91 hours of transcribed audio recordings spoken by two professional speakers . It is the first publicly available large-scale dataset developed to promote Kazakh text-to-speech (TTS) applications in both academia and industry .…

Recently, proposed recurrent neural network (RNN) based all deep learningminimum variance distortionless response (ADL-MVDR) beamformer method yielded superior performance over the conventional MVDR by replacing the matrixinversion and eigenvalue decomposition with two RNNs . Temporal-spatial self-attention module is proposed to better learn thebeamforming weights from the speech and noise spatial spatial spatial covariance matrices .…

The surging usage of electric vehicles (EVs) demand the robust deployment of trustworthy electric vehicle charging station (EVCS) This paper analyses the impact of False Data Injection (FDI) and Distributed Denial of Services (DDoS) attacks on the operation of EVCS .…

A neighborhood restricted Mixed Gibbs Sampling (MGS) based approach is proposed for low-complexity high-order modulation large-scale Multiple-Scale Multiple-InputMultiple-Output (LS-MIMO) detection . Numerical simulation results considering 64-QAM demonstrated that the proposed detectionmethod can substantially improve the MGS algorithm convergence, whereas noextra computational complexity per iteration is required .…

This paper studies scale-free protocol design for H_\infty almost output and regulated output synchronization of heterogeneous multi-agent systems . The collaborative linear protocol designs are based on localizedinformation exchange over the same communication network, which do not require knowledge of the directed network topology and spectrum of the associatedLaplacian matrix .…

Evidence shows that software development methods, frameworks, and even frameworks are seldom applied in companies by following the book . Combinationsof different methodologies into home-grown processes are being constantly covered . In 2016, the HELENA initiative was launched toresearch hybrid development approaches in software system development .…

Modern Code Review (MCR) is a standard in all kinds of organizations thatdevelop software . The goal of this thesis is to boost the efficiency of MCR by developing AI techniques that can partially replace or assist human reviewers . The envisioned techniques distinguish from existing MCR-related AImodels in that we interpret these challenges as graph-learning problems .…

Audio-visual (AV) lip biometrics is a promising authentication technique that combines the benefits of both the audio and visual modalities in speechcommunication . DeepLip outperforms traditional speaker recognition models in context modeling, with an equal error rate of 0.75% and 1.11% on the testdatasets, respectively.…

Approximately 75% of energy used in petrochemical and refining industries is consumed by furnaces . Operating furnaces at optimal conditions results in huge savings . We propose a comprehensive solution to select the best combination of manipulated variables (fired duty, throughput and coil inlet temperature) Satisfying multiple criteria using a cooperative game theory approach .…

Simultaneous wireless information and power transfer (SWIPT) has emerged as aviable technique to energize and connect low-power autonomous devices andenable future Internet of Things (IoT) A major challenge of SWIPT is the energy consumption of the receiver of such low-powered devices .…

Reconfigurable intelligent surfaces (RIS) have drawn considerable attention due to their controllable scattering elements . RISs share somesimilarities with relays, but the two have fundamental differences impacting their performance . A multi-usercommunication system is proposed in this paper wherein a relay and an RIScooperate to improve performance in terms of energy efficiency .…

Many subjective experiments have been performed to develop objective speechintelligibility measures . The novel coronavirus outbreak has made it verydifficult to conduct experiments in a laboratory . One solution is to performremote testing using crowdsourcing . Because we cannot control the listening conditions, it is unclear whether the results are entirely reliable .…

In this paper, we investigate the use of a mass lumped fully explicit timestepping scheme for the discretisation of the wave equation with underlying material parameters that vary at arbitrarily fine scales . We prove that the methodverges with second order in the energy norm, with a leading constant that does not depend on the scales at which the material parameters vary .…

Hercule represents acontract via regulatory norms that capture the involved parties’ expectation . It computes the states of norms (hence, of contracts) from events in the blockchain . Hercule’s novelty and significance lie in that itoperationalizes declarative contracts over semistructured databases .…

Aparametric surface is used to describe the nonplanar road which can describe a combination of curvature, bank and slope . We show that the proposedmodeling approach generalizes planar vehicle models that reference acenterline, such as the Frenet model . We use the proposed approach for vehicle path planning and following using model predictive control .…

We consider the cops and robber game variant consisting of one cop and onerobber on time-varying graphs (TVG) The considered TVGs are edge periodicgraphs, i.e., for each edge, a binary string $s_e$ determines in which timestep the edge is present .…

Chain Reaction Attack exploits weakest point in Online Account Ecosystem . Vulnerability stems from defective multi-factor authentication (MFA), specifically the ones with SMS-based verification, and dependencies among accounts on different platforms . We propose countermeasures including the online account exposedinformation protection mechanism and the built-in authentication to fortify thesecurity of Online Account ecurity of Online Accounts and proposes feasible countermeasures .…

Traditional approaches for leakage detection measure the physical properties of the device and fail to provide root cause analysis . An alternative approach that is gaining traction is to automate leakage detection by modeling the device . We survey the proposed tools to determine the current knowledge levelacross the domain and identify open problems .…

The paper proposes a blockchain-based E2EE framework that can mitigatethe contemporary vulnerabilities in messaging applications . The user’s devicegenerates the public/private key pair during application installation, and asks its mobile network operator (MNO) to issue a digital certificate and store it on the blockchain .…

SSAILFISH outperforms five state-of-the-art smart contract analyzers (SECURITY,MYTHRIL, OYENTE, SEREUM and VANDAL ) in terms of performance, and precision . It discovered 47 previously unknown vulnerable smart contracts outof 89,853 smart contracts from ETHERSCAN . Intotal, SSAFISH discovered 47 unknown .…

Accurately predicting the behavior of a nuclear reactor requires multiphysicssimulation of coupled neutronics, thermal-hydraulics and fuel thermo-mechanics . This work presents methodologies forfull-core surrogate construction based on several realistic equilibrium PWRcore designs . It also sets stage for full-core fuelperformance analysis with BISON where the computational cost becomes moreburdensome.…

Caches have been used to construct various types of covert and side channelsto leak information . Most of the previous cache channels exploit the timingdifference between cache hits and cache misses . We introduce a new andbroader classification of cache covert channel attacks: Hit+Miss, Hit+Hit,Miss+Miss .…

The approach is based on mining Githistories and manipulating a degree of interest (DOI) model . It transforms sourcecode to revitalize feature log levels based on the “interestingness” of the surrounding code . The tool successfully analyzes99.26% of logging statements, increases log level distributions by $20% .…

Automatic program repair (APR) has recently gained attention because it proposes to fix software defects with no human intervention . While APR tools can produce patches that appear to fix the defect for 11-19% of the defects in real-world software, most of the patches produced are not correct or acceptable to developers because they overfit to the tests used during the repair process .…

In this paper, we present an iterative Model Predictive Control (MPC) design for piecewise nonlinear systems . We consider finite time control tasks where the goal of the controller is to steer the system from a starting configuration to a goal state while minimizing a cost function .…

This paper studies the problem of developing an approximate dynamicprogramming (ADP) framework for learning online the value function of aninfinite-horizon optimal problem . We show how these LCBFs can be used to augment alearning-based control policy so as to guarantee safety .…

Recent years have witnessed an upsurge of interest in the problem of anomalydetection on attributed networks . Hop counts based on both global and local contextual information can be served as indicators of anomaly . We propose a hop-countbased model (HCM) to detect anomalies by modeling both local and global contextual information .…

Graph neural networks (GNNs) have been widely used in deep learning ongraphs . They can learn effective node representations that achieve superior performances in graph analysis tasks such as node classification and nodeclustering . However, most methods ignore the heterogeneity in real-worldgraphs and fail to learn complex semantic representations .…

Storm water detention ponds are used to manage the discharge of rainfall runoff from urban areas to nearby streams . Detention ponds are currently designed based on static controls: the output flow of a pond is capped at a fixed value .…

Graph Neural Networks have been widely used for the representationlearning of various structured graph data . RioGNN can learn more discriminative node embedding with enhanced explainability due to therecognition of individual importance of each relation via the filteringthreshold mechanism . A reinforced relation-awareneighbor selection mechanism is developed to choose the most similar neighborsof a targeting node within a relation before aggregating all neighborhood information from different relations to obtain the eventual node embeding .…

An increasing number of studies have sought to understand the processes enacted during software development . These explorations would reveal how team processes occur during all softwaredevelopment efforts . The outcomes suggest that behavioral and intrinsic issues mayinteract with extrinsic factors becoming significant predictors of the speed ofsoftware task resolution .…

Features in product lines and highly configurable systems can interact in ways contrary to developers’ intent . Current methods to identify suchunanticipated feature interactions are costly and inadequate . The contribution of the paper is to use program analysis to extract feature-relevant learning models from the source code in order to detectunwanted feature interactions .…

EAs.LiT is an e-assessment management and analysis software for whichcontextual requirements and usage scenarios changed over time . We consider the microservicearchitecture productive and recommend it for usage in other educational projects . This architectural styleand a few adopted technologies, like RDF as a data format, enabled an eased implementation of various use cases .…