RDMAbox Optimizing RDMA for Memory Intensive Workloads

RDMAbox is a set of low level RDMA opti-mizations that provide better performance than previous ap-proaches . The optimizations are packaged in easy-to-use ker-nel and userspace libraries and presented through simplenodelevel abstractions . The I/O mergequeue at the same time functions as a traffic regulator to enforce admissioncontrol and avoidoverloading the NIC .…

Speeding up Computational Morphogenesis with Online Neural Synthetic Gradients

A wide range of modern science and engineering applications are formulated as optimization problems with partial differential equations (PDEs) These PDE-constrained optimization problems are typically solved using a standard discretize-then-optimize approach . In many industry applicationsthat require high-resolution solutions, the discretized constraints can easilyhave millions or even billions of variables, making it very slow for the standard iterative optimizer to solve the exact gradients .…

The 5th AI City Challenge

The AI City Challenge was created with two goals in mind: pushing the boundaries of research and development in intelligent video analysis for Smarter cities use cases . 305 participating teams across 38 countries, who leveraged city-scale real traffic data and high-quality synthetic data to compete in five challenge tracks .…

A Bi Encoder LSTM Model For Learning Unstructured Dialogs

This paper presents a Long Short Term Memory (LSTM) based architecture that learns unstructured multi-turn dialogs . Ubuntu DialogCorpus Version 2 was used as the corpus for training . We show that our modelachieves 0.8%, 1.0% and 0.3% higher accuracy for Recall@1, Recall@2 and Recall@5 respectively than the benchmark model .…

Prior free Strategic Multiagent Scheduling with focus on Social Distancing

The algorithm takes input from citizens and schedules the store’s time-slots based on their importance to visit the facility . We show that it reduces the socialcongestion significantly using users’ visit data from a store . The problem becomes NP-complete as soon as the multi-slot demands are indivisibleand provide a polynomial-time mechanism that is truthful, individuallyrational, and approximately optimal .…

A Comprehensive Attempt to Research Statement Generation

Research statement generation (RSG) task aims to summarize one’s researchachievements and help prepare a formal research statement . For this task, we construct an RSG dataset with 62 research statements and the corresponding 1,203 publications . We propose a practical RSG method which identifies arearcher’s research directions by topic modeling and clustering techniques .…

Learning Passage Impacts for Inverted Indexes

DeepImpact is a new documentterm-weighting scheme suitable for efficient retrieval using a standardinverted index . Compared to existing methods, it improves impact-scoremodeling and tackles the vocabulary-mismatch problem . When deployed in a re-ranking scenario, it can reach the same effectiveness of state-of-the-art approaches with up to 5.1x speedup inefficiency .…

An Adaptive Learning based Generative Adversarial Network for One To One Voice Conversion

Voice Conversion (VC) deals with conversion of vocal style of one speaker to another speaker while keeping the linguistic contentsunchanged . VC task is performed through a three-stage pipeline consisting of speech analysis, speech feature mapping, and speech reconstruction . ALGAN-VC framework consists of some approaches to improve speech quality and voice similarity between source and target speakers .…

Low rank Tensor Estimation via Riemannian Gauss Newton Statistical Optimality and Second Order Convergence

In this paper, we consider the estimation of a low Tucker rank tensor from anumber of noisy linear measurements . The general problem covers many specificexamples arising from applications, including tensor regression, tensorcompletion, and tensor PCA/SVD . We propose a Riemannian Gauss-Newton (RGN)method with fast implementations for low Tucker Rank tensor estimation .…

A new symmetric linearly implicit exponential integrator preserving polynomial invariants or Lyapunov functions for conservative or dissipative systems

We present a new linearly implicit exponential integrator that preserves thepolynomial first integrals or Lyapunov functions for the conservative and dissipative stiff equations . The method is tested by bothoscillated ordinary differential equations and partial differential equations,e.g., an averaged system in wind-induced oscillation, the Fermi-Pasta-Ulams, and the polynomial pendulum oscillators .…

Superconvergence of Galerkin variational integrators

Galerkin variational integrators approximate avariational (Lagrangian) problem by restricting the space of curves to the set of polynomials of degree at most $s$ and approximating the action integral . We show that, if the quadrature rule is sufficiently accurate, the order of the integrators thus obtained is $2s$.…

Predicting the Number of Reported Bugs in a Software Repository

The bug growth pattern prediction is a complicated, unrelieved task, which needs considerable attention . Advance knowledge of the likely number of bugs discovered in the software system helps software developers in designatingsufficient resources at a convenient time . We observe that LSTM is effective when considering long-runpredictions whereas Random Forest Regressor enriched by exogenous variables performs better for predicting bugs in the short term .…

Fair Capacitated Clustering

Traditionally, clustering algorithms focus on partitioning the data intogroups of similar instances . The similarity objective is notsufficient in applications where a fair-representation of the groups in termsof protected attributes like gender or race is required for each cluster . In many applications, to make the clusters useful for the end-user, abalanced cardinality among the clusters is required .…

Labeling Multipath via Reconfigurable Intelligent Surface

Reconfigurable intelligent surface (RIS) has shown promise in providing apparent benefits in wireless communication and positioning . Each labeled path contains spatial knowledge between the RISand the receiver, thus opening the door for sensing the surrounding world byRISs . The critical challenge is how the labeled paths can be extracted and distinguish from other paths, especially with multipath effects .…

Learning Latent Graph Dynamics for Deformable Object Manipulation

DefOrmable Object Manipulation(G-DOOM) is a long-standing challenge in robotics . It aims to learn latent Graph dynamics for DefOratable Object Manipulations . We train the resulting graph dynamics model through contrastive learning in a high-fidelity simulator . We evaluate a set of challenging cloth and rope manipulation tasks and show that G-Doomperforms a state-of-the-art method .…

Math Operation Embeddings for Open ended Solution Analysis and Feedback

Feedback on student answers and even during intermediate steps in solving questions is an important element in math education . Such feedback can help students correct their errors and ultimately lead toimproved learning outcomes . Most existing approaches for automated studentsolution analysis and feedback require manually constructing cognitive models and anticipating student errors for each question .…

Demystification of Few shot and One shot Learning

Few-shot and one-shot learning have been the subject of active and intensive research in recent years . Theory is based on intrinsic properties of high-dimensional spaces . We show that if the ambient or latent decision space of a learning machine is sufficiently high .dimensional…

Causal Learning for Socially Responsible AI

There have been increasing concerns about Artificial Intelligence due to its unfathomable potential power . Researchers proposed to develop socially responsible AI (SRAI) One of these approaches is causal learning (CL) We survey state-of-the-art methods of CL for SRAI . We begin by examining the seven CL tools to enhance the social responsibility of AI .…

Random Spreading for Unsourced MAC with Power Diversity

We propose an improvement of the random spreading approach with polar codes for unsourced multiple access . Each user encodes its message by a polar code, and the coded bits are then spread using a random spreading sequence . The proposed approach outperforms the existing methods, especially when the number of active users is large, especially in large numbers of users .…

Gaining Insights on Student Course Selection in Higher Education with Community Detection

Gaining insight into course choices holds significant value for universities, especially those who aim for flexibility in their programs and wish to adapt quickly to changing demands of the job market . We found that course choices diversify as programs progress, meaning that attempting to identify a “typical” student gives less insight than understanding what characterizes course choice diversity .…

On the Achievable Sum rate of the RIS aided MIMO Broadcast Channel

Reconfigurable intelligent surfaces (RISs) represent a new technology that can shape the radio wave propagation . We exploit the well-known duality between theGaussian multiple-input multiple- input multiple-output (MIMO) BC and multiple-access channel(MAC) We propose an alternating optimization (AO) algorithm which optimizes the users’ covariance matrices and the RIS phase shifts in the dual MAC .…

Efficient Binary Decision Diagram Manipulation in External Memory

We follow up on the idea of Lars Arge to rephrase the Reduce and Applyalgorithms of Binary Decision Diagrams as iterative I/O-efficient algorithms . These algorithms are implemented in a new BDD library, named Adiar . For instances of about 50 GiB, our algorithms, using external memory, are only upto 3.9 times slower compared to Sylvan, exclusively using internal memory .…