Transfer Learning for Human & AI

The proposed method can be utilized to train agents in environments with fairly complex state and action spaces . The main challenge of using such a rewardfunction is the high sparsity of positive reward signals . To address this problem, we use a simple prediction-based exploration strategy (called CuriousExploration) along with a Return-based Memory Restoration (RMR) technique which tends to remember more valuable memories .…

In standard persistent homology, a persistent cycle born and dying with apersistence interval (bar) associates the bar with a concrete topologicalrepresentative . However, topological features usually go through variations in the lifecycle of a cycle which a single persistent cycle may not capture .…

Modern programming languages (e.g., Java and C#) provide features to separate error-handling code from regular code . The way exception handling code is structured in such languages may lead to multiple, different, and complex control flows, which may affect software testability .…

We consider the weighted completion time minimization problem for capacitatedparallel machines . We bound its approximation ratio with a decreasing function of the ratio between the highest resource demand of any job to the server’s capacity . This research is the first, to the best of ourknowledge, to propose a polynomial-time algorithm with a constant approximationratio for minimizing the weighted sum of job completion times .…

“stack safety” is associated with avariety of compiler, run-time, and hardware mechanisms for protecting stackmemory . We propose a formal characterization of stack safety, formulated with concepts from language-based security . We use these properties to validatethe stack-safety “micro-policies” proposed by Roessler and DeHon [2018].…

An extension to second-order of a new strategy based on in-cell discontinuous reconstructions to deal with this challenging topic is presented . Several numerical tests are proposed to validate the methods for the Coupled-Burgers system, Gasdynamics equations in Lagrangian coordinates and the modified shallow watersystem .…

In this letter, a novel simultaneously transmitting and reflecting (STAR)reconfigurable intelligent surfaces (RISs) design is proposed in anon-orthogonal multiple access (NOMA) enhanced coordinated multi-pointtransmission (CoMP) network . We propose a novelsimultaneously-signal-enhancement-and-cancellation-based (SSECB) design, where the inter-cell interferences and desired signals can be simultaneouslyeliminated and boosted .…

Recently, significant research attention has been devoted to the study ofreconfigurable intelligent surfaces (RISs) RISs are capable of reconfiguring the wireless propagation environment by exploiting the unique properties ofmetamaterials-based integrated large arrays of inexpensive antennas . The medium access control (MAC) of multiple usersaccessing an RIS-enabled channel is still in its infancy, while many open issues remain to be addressed .…

A team of robots that can carefully coordinate their actions can make the entire system robust to individual failures . In order for multi-robot systems to become practical, we need coordinational algorithms that can scale to large teams of robots dealing with dynamicallychanging, failure-prone, contested, and uncertain environments .…

A wireless multihop device-to-device (D2D) basedmobile edge computing (MEC) system, where the destination wireless device (WD) is scheduled to compute nomographic functions . We propose a new multi-level over-the-air (OTA) aggregation scheme for the destination WD tocollect the individual partially aggregated intermediate values (IVAs) for reducing communication resource overhead .…

We present two Dialectica-like constructions for models of intensionalMartin-L\”of type theory . We propose a new semantic notion of finite sum for dependent types, generalizingfinitely-complete extensive categories . The second avoids extensivityassumptions using biproducts in a Kleisli category for a fibred additive monad .…

Survey paper explores six Cloud-based deployment archetypes to achieve highavailability, low end-user latency, and acceptable costs . These are (1) Zonal,(2) Regional, (3) Multi-Regional, (4) Global, (5) Hybrid, and (6) Multi Multi-Clouddeployment archetypes . This enables application owners to better examine the tradeoffsof each deployment model and what is needed for achieving the availability andlatency goals for their application .…

Wireless channels can be inherently privacy-preserving by distorting thereceived signals due to channel noise and superpositioning multiple signalsover-the-air . AirMixML is a differentially private (DP)mechanism limiting the disclosure of each worker’s private sample information at the server, while the worker’s transmit power determines the privacy level .…

Traditionally, graph algorithms get a single graph as input, and then they decide if this graph satisfies a certain property $Phi$ This question is modified in a way that we get a possibly infinite family ofgraphs as an input . We approach this question by using formal languages forspecifying families of graphs .…

We demonstrate that an adversary with a relatively small amount of resources can launch a successful load oscillation attack to destabilize the grid . We evaluated the attack impact using the COSMICtime-domain simulator with two test cases, the IEEE RTS 96 and Polish 2383-BusSystems .…

Large-scale pre-trained models like BERT have obtained a great success invarious Natural Language Processing (NLP) tasks, while it is still a challengeto adapt them to math-related tasks . We propose a novel model, namely MathBERT, which is jointly trained with mathematical formulas and their corresponding contexts .…

The polymer model framework is a classical tool from statistical mechanic . It has recently been used to obtain approximation algorithms for spin systemson classes of bounded-degree graphs . The edge perspective allows us to bound the growthrate of the number of polymers in terms of the total degree of the polymers, which in turn can be related more easily to the expansion properties of the underlying graph .…

Cache-enabled Device-to-Device (D2D) communication is one of the key enablers of the fifth generation (5G) cellular network . But conventional half-duplex(HD) communication may not be sufficient to provide fast enough content delivery over D2D links . In-band full-duple (FD) can provide more content deliveryopportunities, thus resulting improved spectral efficiency and latency reduction .…

In task allocation for real-time domains, alimited number of agents is deployed across a large area to carry out numeroustasks . Agents need to cooperate byforming, disbanding and reforming coalitions . We propose a binary integer program and an anytime and scalableheuristic to solve it .…

Magnetic field sensing can enable fingertracking on a surface with minimal instrumentation . MagSurface allows more flexibility in how to interact with everyday surfaces . Localization error as low as a few millimeters was achieved in tests of the system .…

Existing methods of medical visual question answering usually rely on transfer learning to obtain effective image feature representation . We introduce a cross-modal self-attention module to capture the long-range contextual relevance for more effectivefusion of visual and linguistic features . Experimental results demonstrate that the proposed method outperforms existing state-of-the-art methods .…

Bayesian Knowledge Tracing, a model used for cognitive mastery estimation, has been a hallmark of adaptive learning research and an integral component of intelligent tutoring systems . PyBKT is an accessible and computationally efficient library of model extensions from the literature .…

BI-REC is a conversational recommendation system for BI applications to help users accomplish their data analysis tasks . We define the space of data analysis in terms of BI patterns, augmented with rich semantic information extracted from the OLAP cube definition, and usegraph embeddings learned using GraphSAGE to create a compact representation of the analysis state .…

Rate-Splitting Multiple Access (RSMA) is a flexible and robust multipleaccess scheme for downlink multi-antenna wireless networks . RSMA relies onRate-splitting (RS) at the transmitter and Successive Interference Cancellation (SIC) at receivers . We present the optimal system designsemploying RSMA that target short-packet and low-latency communications as well as robust communications with high-throughput under the practical and importantsetup of imperfect Channel State Information at Transmitter (CSIT) originating from user mobility and feedback latency in the network .…

Physical-layer key generation (PKG) can generate symmetric keys between twocommunication ends based on the reciprocal uplink and downlink channels . Bysmartly reconfiguring the radio signal propagation, intelligent reflectingsurface (IRS) is able to improve the secret key rate of PKG . IRS-assistedmultiple-input single-output (MISO) system aims to maximize the secretkey rate by optimally designing the IRS passive beamforming .…

The most commonly used interface between a video game and a human user is a “game controller” We present simple formats for such mappings as well as constraints on possible inputs which are determined by a physical controller or required to be met for a game software, along with methodsto transform said constraints via a button-action mapping and to check oneconstraint set against another, i.e.…

EnsembleConvolutional Neural Network (ECNN) is trained using a Mixture of Experts (MOE) model that integrates opinions and determines the final grasping decision . ECNN introduces minimal computational cost overhead, and thenetwork can virtually run as fast as the slowest expert .…

Lectures were developed for the course Computational SocialChoice of the Artificial Intelligence MSc programme at the University of Groningen . They cover mathematical and algorithmic aspects of voting theory .…

The Multi-Decoder model for speech translation outperforms the previous state-of-the-art by around +6 and +3 BLEU on Fisher-CallHome and MuST-C . The model can also incorporate external models at intermediate stages of the network to re-score or adapt towards out-of the domain data .…

Data mining project managers can benefit from using standard data miningprocess models . The CRISP-IM facilitates idea generation, through theuse of Dynamic Topic Modeling (DTM) and unsupervised machine learning, and statistical analysis on a dataset of scholarly articles. The adaptedCRISp-IM can be used to guide the process of identifying trends using scholarlyliterature datasets or temporally organized patent or any other textual dataset to elicit ideas.…

ParadisEO is a comprehensive C++ freesoftware which targets the development of modular metaheuristics . It is of prior importance to have access to mature and flexible software frameworks which allow for an efficient exploration of the algorithm design space . This articlesummarizes the features of the framework, a highly modular architecture, a large set of components, speed of execution and automated algorithm design features, which are key to modern approaches to metaheiristics development .…

We propose a data-driven model order reduction (MOR) technique for partial differential equations that exhibit parameter-dependentjump-discontinuities . We build upon themethodology of approximating the map between the parameter domain and theexpansion coefficients of the reduced basis via regression . The online stagequeries the regression model for the expansion coefficients and recovers areduced approximation for the solution .…

Myntra is an online fashion e-commerce company based in India . Ouralgorithm to identify eligible/elite customers is based on sound principles ofgame theory . It is simple, easy to implement and scalable. Oural algorithm to identify eligibility/elitation customers is simple and easy to use.…

Recent years have witnessed an increasing interest in improving theperception performance of LiDARs on autonomous vehicles . We propose an easy-to-computeinformation-theoretic surrogate cost metric based on Probabilistic OccupancyGrids . We show acorrelation between our surrogate function and common object detectionperformance metrics .…

Massive multiple- input multiple-input multiple-output (MIMO) is one of the key techniques toachieve better spectrum and energy efficiency in 5G system . The channel stateinformation (CSI) needs to be fed back from the user equipment to the basestation in frequency division duplexing (FDD) mode .…

Fixed point construction is a method for designing tile sets and cellularautomata with highly nontrivial dynamical and computational properties . It produces an infinite hierarchy of systems where each layer simulates the nextone . The simulations are implemented entirely by computations of Turingmachines embedded in the tilings or spacetime diagrams .…

Graph partitioning has been an important tool to partition the work among processors to minimize the communication cost and balance the workload . Sphynx provides a good and robust partitioning method across a wide range of graphs for applications looking for a GPU-based partitioner .…

Sequential Recommendation characterizes the evolving patterns by modelingitem sequences chronologically . We use thistransformer to generate fabricated historical items at the beginning of short sequences . We fine-tune the transformer using these augmentedsequences from the time order to predict the next item .…

Alice and Bob take turns on choosing consecutive letters of a word, with Alice winning if the result lies in a predetermined language . The turn orders that result in a win for Alice form a binary language that is regular whenever the target language is, and we prove someupper and lower bounds for its state complexity based on that of the targetlanguage .…

In 1983, G. Ramharter gave an explicit description of the extremal arrangements of the regular continued fraction and theminimizing arrangement for the semi-regular continued fraction . He showed that if $|A|=2$ then the maximizing arrangement is unique(up to reversal) and depends only on the partition $P$ and not on the values of the digits in $A$ .…

The study of network robustness is a critical tool in the characterizationand sense making of complex interconnected systems such as infrastructure,communication and social networks . This survey guides researchers and practitioners in navigating theexpansive field, while summarizing answers to key questions .…

The proposed approach targets to gradually correct themislabelled data samples during training in a self-supervised fashion, without any prior assumption on the amount of label noise . We benchmark our approach on several time-series classification datasets and find it to be comparable and sometimes better than state-of-the-art methods .…

There is an opportunity for deep learning to revolutionize science and technology by revealing its findings in a human interpretable manner . Wedevelop a novel data-driven approach for creating a human-machine partnershipto accelerate scientific discovery . By collecting physical system responses,under carefully selected excitations, we train rational neural networks tolearn Green’s functions of hidden partial differential equation .…

A real-time system, we specify the temporal properties on QCTMC by signal temporal logic (STL) To effectivelycheck the atomic propositions in STL, we develop a state-of-art real rootisolation algorithm under Schanuel’s conjecture . Further, we check the generalSTL formula by interval operations with a bottom-up fashion, whose querycomplexity turns out to be linear in the size of the input formula by calling the real root isolation algorithm .…

Distribution matchers for finite alphabets are shown to have informationaldivergences that grow logarithmically with the block length . This is a generalizing abasic result for binary strings .…

In empirical software engineering, benchmarks can be used for comparing methods, techniques and tools . The recent ACM SIGSOFTEmpirical Standards for Software Engineering Research do not include an explicit checklist for benchmarking . In this paper, we discuss benchmarks forsoftware performance and scalability evaluation as example research areas in software engineering .…

The age of the Internet of Things is adding an increasing number of new devices to the Internet and is expected to have fifty billion connected units by 2021 . The review identifies nineteen areas of vulnerability in the LoRaWAN protocol .…

Two larger multilingual masked language models with 3.5B and 10.7B parameters outperform XLM-R by 1.8% and 2.4% average accuracy on XNLI . Our model also outperforms theRoBERTa-Large model on several English tasks of the GLUE benchmark by 0.3% onaverage while handling 99 more languages .…

In recent years, we have witnessed a new kind of DDoS attack, the burstattack . The attacker launches periodic bursts of traffic overload on online targets . The auto-scaling mechanism translates the flat DDoS attacks into Economic Denial of Sustainability attacks .…