The Communication Complexity of Payment Computation

In many important settings, $f$ uniquely determines $P$ (up to a constant) A common approach is to focus on the design of an incentive compatible mechanism and neglect the role of the payment function . Fadel and Segal show that for every $f$, $cc_{IC}(f) is strictly harder than computing only the output .…

Autocratic Strategies of Multi State Games

In a single-state repeated game, zero-determinant strategies can unilaterally force functions of the payoffs to take values in particular closed intervals . When the explicit use of a determinant is absent from the analysis, they are instead called autocratic . We extend their study to the setting of finite stategames with deterministic transitions .…

Can You be More Social Injecting Politeness and Positivity into Task Oriented Conversational Agents

Goal-oriented conversational agents are becoming prevalent in our daily lives . For these systems to engage users and achieve their goals, they need to exhibit appropriate social behavior as well as provide informative replies that guide users through tasks . Analyses show that social languageused by human agents is associated with greater users’ responsiveness and taskcompletion .…

Causal Perception in Question Answering Systems

Using question-answering systems often produce questionable causal claims . We found that in a system that occasionally providedreasonable responses, showing a scatterplot increased the plausibility of such claims . Simply warning participants that correlationis not causation seemed to lead participants to accept reasonable causal claims more cautiously .…

Meta Adaptive Neural Ranking with Contrastive Synthetic Supervision

Neu-IR models have shown their effectivenessand thrive from end-to-end training with massive high-quality relevance labels . But relevance labels at such quantity are luxury and unavailable in many ranking scenarios, for example, in biomedical search . We first leverage contrastive querygeneration (ContrastQG) to synthesize more informative queries as in-domainweak relevance labels, and then filter them with meta adaptive learning to rank(MetaLTR) to better generalize neural rankers to the target few-shot domain .…

Supporting Human Memory by Reconstructing Personal Episodic Narratives from Digital Traces

Personal Digital Traces – PDTs – can be used to reconstruct people’s episodic memories and connect to their past personalevents . This reconstruction has several applications, from helping patients with neurodegenerative diseases recall past events to gathering clues from multiple sources to identify recent contacts and places visited – a critical new application for the current health crisis .…

Abstractive Query Focused Summarization with Query Free Resources

The availability of large-scale datasets has driven the development of neuralsequence-to-sequence models to generate generic summaries . Despite learning from minimal supervision, our system achieves state-of-the-art results in the distantly supervised setting across domains and query types . To further utilize generic data forgeneration, three attributes are incorporated during training and inference to control the shape of the final summary: evidence rank, query guidance, andsummary length .…

Hybrid Interest Modeling for Long tailed Users

User behavior modeling is a key technique for recommender systems . Most methods focus on head users with large-scale interactions and hence suffer from data sparsity issues . We argue that current methods are limited by the strict privacy policy and have low scalability in real-world applications .…

Image to Image Retrieval by Learning Similarity between Scene Graphs

The similarity between scene graphs of two images reflects the relevance of their contents . Graph neural networks are trained to predict the proxy image relevance measure, computed from captions using a pre-trained sentence similarity model . The collected dataset shows thatour method agrees well with the human perception of image similarity than othercompetitive baselines .…

A Number Theoretic Approach for Fast Discovery of Single Hop Wireless Networks

Interference management has become a key factor in regulating transmissions in wireless communication networks . To support effective interferencemanagement schemes, it can be essential to have prior knowledge about thenetwork topology . In this paper, we build on existing results in the literature on the simulation of the message passing model, and present an efficient strategy for fast discovery of the network topology during a pilotcommunication phase .…

A Priori Generalization Analysis of the Deep Ritz Method for Solving High Dimensional Elliptic Equations

This paper concerns the a priori generalization analysis of the Deep RitzMethod [W. E and B. Yu, 2017], a popular neural-network-based method forsolving high dimensional partial differential equations . We derive the generalization error bounds of two-layer neural networks in the framework of the DRM for solving two prototype elliptic PDEs: Poisson equation and staticSchrodinger equation on the $d-dimensional unit hypercube .…

Explainable Reactive Synthesis

Reactive synthesis transforms a specification of a reactive system into an implementation . The main disadvantage is that the implementation is difficult to understand . We present SAT-based algorithms for thesynthesis of repairs and explanations . The algorithms are evaluated on a range of examples including benchmarks taken from the SYNTCOMP competition .…

Prosocial Norm Emergence in Multiagent Systems

The social structure of a multiagentsystem can be reflected in social norms among its members . We propose Cha, a framework for the emergence of prosocialnorms . Cha agents incorporate prosocialdecision making based on inequity aversion theory, reflecting an intuition of guilt from being antisocial .…

Emergent Symbols through Binding in External Memory

Deep neural network algorithms have proven to be a powerful tool for learning directly from high-dimensional data, but currently lack this capacity for data-efficient induction of abstract rules . The Emergent Symbol Binding Network (ESBN) is a recurrent network with an external memory that enables a form of variable-binding andindirection .…

Joint Beamforming and Power Control for Throughput Maximization in IRS assisted MISO WPCNs

Intelligent reflecting surface (IRS) is an emerging technology to enhance the energy- and spectrum-efficiency of wireless powered communication networks . In this paper, we investigate an IRS-assisted multiuser multiple-inputsingle-output (MISO) WPCN . We jointly optimize the active beamforming of the HAP and the reflecting coefficients (passive beamforming) of the IRS in both DL and UL transmissions .…

Sensifi A Wireless Sensing System for Ultra High Rate Applications

Wireless Sensor Networks (WSNs) are being used in various applications such as structural health monitoring and industrial control . The existing WSNs primarily rely on low-power, low-rate wireless technologies such as 802.15.4 and Bluetooth . In this paper, we strive to tackle the challenges of developing ultra-high-rateWSNS based on 802.11 (WiFi) standard by proposing Sensifi.…

Alternative Paths Planner APP for Provably Fixed time Manipulation Planning in Semi structured Environments

Alternative Paths Planner (APP) provides fixed-time planning guarantees in semi-structured environments . APPplans a set of alternative paths offline such that, for any configuration of the movable obstacles, at least one of the paths from this set is collision-free . APP is several orders of magnitude faster than state-of-the-art motion planners foreach domain .…

Detection of Lexical Stress Errors in Non native L2 English with Data Augmentation and Attention

This paper describes two novel complementary techniques that improve the detection of lexical stress errors in non-native (L2) English speech . In a classical approach, audio features are usually extracted from fixed regions of speech such as syllable nucleus . We propose anattention-based deep learning model that automatically derives optimalsyllable-level representation from frame-level and phoneme-level audiofeatures .…

Data driven audio recognition a supervised dictionary approach

Machine hearing is an emerging area . We propose ageneric and data-driven representation learning approach . Method is capable to reach state-of-the-art hand-crafted features forboth applications . It is anovel and efficient supervised dictionary learning method is presented . The results show that our method is capable of reaching state of the art handcrafted features .…

Counting the Number of Solutions to Constraints

In this paper, we survey research works on the problems of counting the number of solutions to constraints . The constraints may take various forms, including formulas in the propositional logic, linearinequalities over the reals or integers . We describe some techniques and tools for solving the counting problems, as well as some applications (e.g.,…

GAKP GRU Association and Kalman Prediction for Multiple Object Tracking

Multiple Object Tracking (MOT) has been a useful yet challenging task in manyreal-world applications such as video surveillance, intelligent retail, and smart city . The challenge is how to model long-term temporal dependencies in anefficient manner . Some recent works employ Recurrent Neural Networks (RNN) to obtain good performance, which, however, requires a large amount of training data .…

Longitudinal diffusion MRI analysis using Segis Net a single step deep learning framework for simultaneous segmentation and registration

This work presents a single-step deep-learning framework for longitudinalimage analysis, coined Segis-Net . To optimally exploit information available in longitudinal data, this method concurrently learns a multi-class segmentationand nonlinear registration . Segmentation and registration are modeled using aconvolutional neural network and optimized simultaneously for their mutualbenefit .…

Red Dragon AI at TextGraphs 2020 Shared Task LIT LSTM Interleaved Transformer for Multi Hop Explanation Ranking

The LSTM-Interleaved Transformer incorporates cross-document interactions for improved multi-hop ranking . The LITarchitecture can leverage prior ranking positions in the re-ranking setting . The model is competitive on the current leaderboard for the TextGraphs 2020 task, achieving a test-set MAP of 0.5607, and would have gained third place had we submitted before the competition deadline .…

Federated Multi Agent Actor Critic Learning for Age Sensitive Mobile Edge Computing

Mobile edge computing (MEC) introduces a new scheme for various distributed communication-computing systems such as industrial Internet of Things (IoT), vehicular communication, smart city, etc. As an emerging technique, mobile edge computing introduces new schemes for various systems . A novel policybased multi-agent deep reinforcement learning (RL) framework is proposed as a paradigm for joint collaboration in the investigated MEC systems, where edge devices and center controller learn the interactive strategies through their own observations .…

Automatic Curriculum Learning With Over repetition Penalty for Dialogue Policy Learning

Dialogue policy learning based on reinforcement learning is difficult to be applied to real users to train dialogue agents from scratch . The ACL-DQN significantly improves the effectiveness and stability of dialoguetasks with a statistically significant margin . The teacher model arranges a meaningful ordered curriculum and automatically adjusts it by monitoring the learning progress of the dialogue agent and the over-repetition penalty without any requirement of prior knowledge .…