Tong Zhang

Towards More Efficient SPSD Matrix Approximation and CUR Matrix Decomposition

2015 · arXiv (Cornell University)

Symmetric positive semi-definite (SPSD) matrix approximation methods have been extensively used to speed up large-scale eigenvalue computation and kernel learning methods. The standard sketch based method, which we call the prototype model, produces relatively accurate …

Application of frequent item set mining algorithm in IDS based on Hadoop framework

2018

With the coming of big data time, the huge number of IDS log makes the traditional computing technology and systems cannot cope and deal with the needs of the analysis of security log, so large-scale …

Active and Adaptive Application-Level Flow Control for Latency Sensitive RPC Applications

2019

The Remote Procedure Call (RPC) frameworks are widely deployed in industry. Applications supported by RPC frameworks are often latency-sensitive which strictly require to be responded before the deadline. For meeting this requirement, RPC frameworks adopt …

Disentangled Generative Causal Representation Learning

2021

This paper proposes a Disentangled gEnerative cAusal Representation (DEAR) learning method. Unlike existing disentanglement methods that enforce independence of the latent variables, we consider the general case where the underlying factors of interests can be …

WAL-assisted Tiering: Painlessly Improving Your Favorite Log-Structured KV Store Instead of Building a New One

2020

This paper presents a simple design approach that can be easily integrated into existing mature log-structured key-value (KV) stores (e.g., RocksDB) to mitigate the impact of background compaction. Reducing compaction-induced performance degradation has been widely …

Universal Adder Neural Networks

2021 · arXiv (Cornell University)

Compared with cheap addition operation, multiplication operation is of much higher computation complexity. The widely-used convolutions in deep neural networks are exactly cross-correlation to measure the similarity between input feature and convolution filters, which involves …

A Theoretical Analysis on Independence-driven Importance Weighting for Covariate-shift Generalization

2021 · arXiv (Cornell University)

Covariate-shift generalization, a typical case in out-of-distribution (OOD) generalization, requires a good performance on the unknown test distribution, which varies from the accessible training distribution in the form of covariate shift. Recently, independence-driven importance weighting …

A Feasibility Study of Low-frequency Ultrasound Tomography for Human Thorax

2022 · 2022 IEEE International Ultrasonics Symposium (IUS)

Ultrasound is widely used in biomedical imaging. High-frequency ultrasound at MHz is usually used for good resolution. However, ultrasound at this band cannot permeate human thorax, for it is strongly scattered and reflected by air …

Automatic Prompt Augmentation and Selection with Chain-of-Thought from Labeled Data

2023

Chain-of-thought (CoT) advances the reasoning abilities of large language models (LLMs) and achieves superior performance in complex reasoning tasks. However, most CoT studies rely on carefully designed human-annotated rational chains to prompt LLMs, posing challenges …

MLLM-Protector: Ensuring MLLM's Safety without Hurting Performance

2024 · arXiv (Cornell University)

The deployment of multimodal large language models (MLLMs) has brought forth a unique vulnerability: susceptibility to malicious attacks through visual inputs. This paper investigates the novel challenge of defending MLLMs against such attacks. Compared to …

Wasserstein Discriminant Dictionary Learning for Graph Representation

2024 · IEEE Transactions on Pattern Analysis and Machine Intelligence

Mining discriminative graph topological information plays an important role in promoting graph representation ability. However, it suffers from two main issues: (1) the difficulty/complexity of computing global inter-class/intra-class scatters, commonly related to mean and covariance …

A New Perspective on Time Series Anomaly Detection: Faster Patch-based Broad Learning System

2024 · arXiv (Cornell University)

Time series anomaly detection (TSAD) has been a research hotspot in both academia and industry in recent years. Deep learning methods have become the mainstream research direction due to their excellent performance. However, new viewpoints …

Adaptive Sampling Towards Fast Graph Representation Learning

2018 · arXiv (Cornell University)

Graph Convolutional Networks (GCNs) have become a crucial tool on learning representations of graph vertices. The main challenge of adapting GCNs on large-scale graphs is the scalability issue that it incurs heavy cost both in …

Multi-Head Attention with Disagreement Regularization

2018

Multi-head attention is appealing for the ability to jointly attend to information from different representation subspaces at different positions. In this work, we introduce a disagreement regularization to explicitly encourage the diversity among multiple attention …

Learning to Remember Translation History with a Continuous Cache

2018 · Transactions of the Association for Computational Linguistics

Existing neural machine translation (NMT) models generally translate sentences in isolation, missing the opportunity to take advantage of document-level information. In this work, we propose to augment NMT models with a very light-weight cache-like memory …

Effective Use of Word Order for Text Categorization with Convolutional Neural Networks

2015

Convolutional neural network (CNN) is a neural network that can make use of the internal structure of data such as the 2D structure of image data. This paper studies CNN on text categorization to exploit …

Modeling Localness for Self-Attention Networks

2018

Self-attention networks have proven to be of profound value for its strength of capturing global dependencies. In this work, we propose to model localness for self-attention networks, which enhances the ability of capturing useful local …

ZEN: Pre-training Chinese Text Encoder Enhanced by N-gram Representations

2020

The pre-training of text encoders normally processes text as a sequence of tokens corresponding to small text units, such as word pieces in English and characters in Chinese. It omits information carried by larger text …