Faster Algorithms for Bounded Difference Min Plus Product

Min-plus product of two $n\times n$ matrices is a fundamental problem in algorithm research . It is known to be equivalent to APSP, and in general it hasno truly subcubic algorithms . Our algorithm runs in randomized time $O(n^{2.779) by the fastrectangular matrix multiplication algorithm [Le Gall \& Urrutia 18], better than $2.791$ ($\omega <2.373)$ [Alman \&V.V.Williams 20] …

Multifractal of mass function

Multifractal plays an important role in many fields, but there is fewattentions about mass function, which can better deal with uncertain information than probability . In this paper, we proposed multifractal of massfunction . One interesting property is that the multifractable dimensionof mass function with maximum entropy is 1.585 no matter the order .…

Learning First Order Rules with Relational Path Contrast for Inductive Relation Reasoning

Relation reasoning in knowledge graphs aims at predicting missingrelations in incomplete triples . The dominant paradigm is learning theembeddings of relations and entities, which is limited to a transductivesetting . Previous inductive methods are scalable and consume less resource . We propose a novel graph convolutional network(GCN)-based approach for interpretable inductive reasoning with relational pathcontrast, named RPC-IR .…

Multimodal Dialogue Response Generation

Amultimodal dialogue generation model takes the dialogue history as input, then generates a textual sequence or an image as response . Learning sucha model often requires multimodal dialogues containing both texts and images which are difficult to obtain . The method achieves state-of-the-art results in bothautomatic and human evaluation, and can generate informative text and high-resolution image responses.…

Perturbative construction of mean field equations in extensive rank matrix factorization and denoising

Factorization of matrices where the rank of the two factors diverges linearly with their sizes has many applications in diverse areas such as unsupervisedrepresentation learning, dictionary learning or sparse coding . In the limit where thedimensions of the matrices tend to infinity, but their ratios remain fixed, we expect to be able to derive closed form expressions for the optimal meansquared error on the estimation of two factors .…

Measuring Total Transverse Reference free Displacements of Railroad Bridges using 2 Degrees of Freedom 2DOF Experimental Validation

Railroad bridge engineers are interested in the displacement of railroadbridges when the train is crossing the bridge for engineering decision making . Measuring displacements under train crossing events is difficult . If simplified reference-free methods would be accurate andvalidated, owners would conduct objective performance assessment of their bridge inventories under trains .…

Challenges Porting a C Template Metaprogramming Abstraction Layer to Directive based Offloading

HPC systems employ a growing variety of compute accelerators with differentarchitectures and from different vendors . Large scientific applications need to run efficiently across these systems but need to retain a singlecode-base . Directive-based offloading models set out to provide the required portability, but, to existing codes, they themselves represent yet another API to port to .…

Learning Continuous Chaotic Attractors with a Reservoir Computer

The exact mechanisms of how dynamical neural systems performabstraction are still not well-understood . We train a 1000-neuron RNN — a reservoircomputer — to abstract a continuous dynamical attractor memory . We propose a theoretical mechanism of this abstractionby combining ideas from differentiable generalized synchronization and feedbackdynamics .…

An LSTM based Plagiarism Detection via Attention Mechanism and a Population based Approach for Pre Training Parameters with imbalanced Classes

This paper proposes an architecture based on a Long Short-TermMemory (LSTM) and attention mechanism called LSTM-AM-ABC boosted by apopulation-based approach for parameter initialization . The results clearly show that the proposed method can provide competitive performance . Our proposed algorithm can find the initial values for model learning in all LSTm, attentionmechanism, and feed-forward neural network, simultaneously.…

Making Existing Software Quantum Safe Lessons Learned

In the era of quantum computing, Shor’s algorithm running on quantumcomputers (QCs) can break asymmetric encryption algorithms that classicalcomputers essentially cannot . QCs, with the help of Grover’s algorithm, can speed up the breaking of symmetric encryption . The exact date when QCs will become “dangerous” for practical problems is unknown, the consensus is that this future is near .…

ACOA Chronological Analysis of the Exhibition of Artistic Works

The ACOA platform supports the organization of multiple sources of information related to creative processes behind complex works . This information is of great interest to conservators and curators, as well as to the general public . This platform houses achronological evolution of the work, through the contextual dissemination of multimedia content .…

Verification of MPI programs

In this paper, we outline an approach to verifying parallel programs . A newmathematical model of parallel programs is introduced . The introduced model is illustrated by the verification of the matrix multiplication MPI program .…

Evolving spiking neuron cellular automata and networks to emulate in vitro neuronal activity

Neuro-inspired models and systems have great potential for applications inunconventional computing . Often, the mechanisms of biological neurons aremodeled or mimicked in simulated or physical systems in an attempt to harness some of the computational power of the brain . The aim of this approach was to develop a method of producing models capable of exhibiting complex behavior that may be suitable for use as computationalsubstrates .…

Self stabilizing Byzantine and Intrusion tolerant Consensus

Consensus applications include services for the Cloud or Blockchain . We propose the first self-stabilizing solution for intrusion-tolerant multivalued consensus for message-passing systems prone to Byzantine failures . This work aims at the design of an even more robust solution than MR, which can deal with up to $t < n/3$ Byzantine processes, where $n$ is the number of processes . In addition totolerating Byzantine and communication failures, self-Stabilizing systems canautomatically recover after the occurrence of \emph{arbitrarytransient-faults}. These faults represent any violation of the assumptions according to which the system was designed to operate (provided that the algorithm code remains intact). We propose, to the best of our knowledge, the first such a solution to such a fault-tolerance solution to this type of problem . Back to Mailing Back to Back to the page you came from :http://www.mailing-backto the page.com/news/post/daily/mailing back to page/back to page to-the-page to the mainline . …

A Heterogeneous Graph Based Framework for Multimodal Neuroimaging Fusion Learning

Traditional GNN-based models usually assume the brain network is a homogeneous graph with single type of nodes and edges . We present a Heterogeneous Graph neural network for Multimodalneuroimaging fusion learning (HGM) We also propose a self-supervisedpre-training strategy based on heterogeneou brain network to address theoverfitting problem due to the complex model and small sample size .…

Generalized Out of Distribution Detection A Survey

Out-of-distribution (OOD) detection is critical to ensuring the reliabilityand safety of machine learning systems . Other problems related to OOD detection include anomaly detection (AD), novelty detection (ND), open set recognition (OSR), and outlier detection (OD) Despite having different definitions and settings, these problems often confuse readers and practitioners, some existing studies misuse terms .…

Voting Theory in the Lean Theorem Prover

In this paper, we report on the development of a framework for formalizing voting theory in the Lean theorem prover . In order to formalize voting theoretic axioms concerning addingor removing candidates and voters, we work in a variable-election setting whoseformalization makes use of dependent types in Lean .…

A Variational Bayesian Approach to Learning Latent Variables for Acoustic Knowledge Transfer

We propose a variational Bayesian (VB) approach to learning distributions of latent variables in deep neural network (DNN) models for cross-domain knowledgetransfer . To accomplish model transfer, knowledge learnt from a source domainis encoded in prior distributions of . latent variables and optimally combined,in a Bayesian sense, with a small set of adaptation data from a target domain to approximate the corresponding posterior distributions .…

Sum of Squares Geometry Processing

Geometry processing presents a variety of difficult numerical problems . Sum-of-squares optimization transforms nonlinear polynomial optimization problems into convex problems whose complexity is captured by a single degreeparameter . This allows us to solve a suite of problems on higher-ordersurfaces, such as continuous collision detection and closest point queries oncurved patches, with only minor changes between formulations and geometries .…

Directional forces in the evolution of grammar

In modern English grammar, the perfect is formed with have+PP (pastparticiple), but in older English the be+PP form existed as well . This finding strongly suggests that the perfect could have evolved through natural selection rather than random drift . Most intransitive verbs exhibited an increase in the frequency of have +PP, some of which passed the Frequency Increment Test (FIT), indicating a rapid S-shapeincrease .…

Physical Side Channel Attacks on Embedded Neural Networks A Survey

Deep Neural Networks (DNN) have progressively beenintegrated on all types of platforms, from data centers to embedded systems including low-power processors and, recently, FPGAs . The underlying hardware securityvulnerabilities of embedded NN implementations remain unaddressed . Inparticular, embedded DNN implementations are vulnerable to Side-ChannelAnalysis (SCA) attacks, which are especially important in the IoT and edgecomputing contexts where an attacker can usually gain physical access to thetargeted device .…

Physics guided Deep Markov Models for Learning Nonlinear Dynamical Systems with Uncertainty

In this paper, we propose a probabilistic physics-guided framework, termedPhysics-guided Deep Markov Model . The framework is especially targetedto the inference of the characteristics and latent structure of nonlineardynamical systems from measurement data . The proposed framework takes advantage of the expressive power of deep learning, while retaining the driving physics of the dynamical system by imposing physics-driven restrictions on the side of the latent space .…

Verification of MPI programs

In this paper, we outline an approach to verifying parallel programs . A newmathematical model of parallel programs is introduced . The introduced model is illustrated by the verification of the matrix multiplication MPI program .…

Collaborating with Humans without Human Data

Collaborating with humans requires rapidly adapting to their individual strengths, weaknesses, and preferences . Most standardmulti-agent reinforcement learning techniques produce agents that overfit to their training partners . Alternatively, researchers can collecthuman data, train a human model using behavioral cloning, and then use thatmodel to train “human-aware” agents .…

Streaming Decision Trees and Forests

Machine learning has successfully leveraged modern data and provided solutions to innumerable real-world problems . Currently, estimators could handle both scenarios with all samples available and situations requiring continuous updates . However, there is still room for improvement on streaming algorithms based onbatch decision trees and random forests, which are the leading methods in batch data tasks .…