Search Results for author:
Found 49 papers, 4 papers with code
InfLoRA: Interference-Free Low-Rank Adaptation for Continual Learning
Furthermore, InfLoRA designs this subspace to eliminate the interference of the new task on the old tasks, making a good trade-off between stability and plasticity.
UniAP: Unifying Inter- and Intra-Layer Automatic Parallelism by Mixed Integer Quadratic Programming
To the best of our knowledge, UniAP is the first parallel method that can jointly optimize the two categories of parallel strategies to find an optimal solution.
On the Optimal Batch Size for Byzantine-Robust Distributed Learning
However, for existing BRDL methods, large batch sizes will lead to a drop on model accuracy, even if there is no Byzantine attack.
GNNFormer: A Graph-based Framework for Cytopathology Report Generation
A common weakness of these works is that they do not explicitly model the structural information among cells, which is a key feature of pathology images and provides significant information for making diagnoses.
FedREP: A Byzantine-Robust, Communication-Efficient and Privacy-Preserving Framework for Federated Learning
Based on ConSpar, we further propose a novel FL framework called FedREP, which is Byzantine-robust, communication-efficient and privacy-preserving.
Asymmetric Learning for Graph Neural Network based Link Prediction
To the best of our knowledge, AML is the first GNN-LP method adopting an asymmetric learning strategy for node representation learning.
Adaptive Plasticity Improvement for Continual Learning
Besides the ability to overcome CF on old tasks, API also tries to evaluate the model's plasticity and then adaptively improve the model's plasticity for learning a new task if necessary.
Optimizing Class Distribution in Memory for Multi-Label Online Continual Learning
But class distribution in memory is critical for replay-based memory to get good performance, especially when the class distribution in data stream is highly imbalanced.
State-based Episodic Memory for Multi-Agent Reinforcement Learning
SEM adopts episodic memory (EM) to supervise the centralized training procedure of CTDE in MARL.
Full-Precision Free Binary Graph Neural Networks
Binary neural networks have become a promising research topic due to their fast inference speed and low energy consumption advantages.
Optimizing Class Distribution in Memory for Multi-Label Continual Learning
Most existing replay-based methods focus on single-label problems in which each sample in the data stream has only one label.
Cross Domain Robot Imitation with Invariant Representation
In this paper, we consider CDIL on a class of similar robots.
Multiple Code Hashing for Efficient Image Retrieval
In this paper, we propose a novel hashing framework, called multiple code hashing (MCH), to improve the performance of hash bucket search.
ExchNet: A Unified Hashing Network for Large-Scale Fine-Grained Image Retrieval
Retrieving content relevant images from a large-scale fine-grained dataset could suffer from intolerably slow query speed and highly redundant storage cost, due to high-dimensional real-valued embeddings which aim to distinguish subtle visual differences of fine-grained objects.
Stochastic Normalized Gradient Descent with Momentum for Large-Batch Training
Empirical results on deep learning verify that when adopting the same large batch size, SNGM can achieve better test accuracy than MSGD and other state-of-the-art large-batch training methods.
TOMA: Topological Map Abstraction for Reinforcement Learning
In particular, we propose planning to explore, in which TOMA is used to accelerate exploration by guiding the agent towards unexplored states.
Buffered Asynchronous SGD for Byzantine Learning
In this paper, we propose a novel method, called buffered asynchronous stochastic gradient descent (BASGD), for ABL.
Stagewise Enlargement of Batch Size for SGD-based Learning
We theoretically prove that, compared to classical stagewise SGD which decreases learning rate by stage, \mbox{SEBS} can reduce the number of parameter updates without increasing generalization error.
Weight Normalization based Quantization for Deep Neural Network Compression
WNQ adopts weight normalization to avoid the long-tail distribution of network weights and subsequently reduces the quantization error.
ADASS: Adaptive Sample Selection for Training Acceleration
However, in all existing SGD and its variants, the sample size in each iteration~(epoch) of training is the same as the size of the full training set.
Clustered Reinforcement Learning
Exploration strategy design is one of the challenging problems in reinforcement learning~(RL), especially when the environment contains a large state space or sparse rewards.
Global Momentum Compression for Sparse Communication in Distributed Learning
With the rapid growth of data, distributed momentum stochastic gradient descent~(DMSGD) has been widely used in distributed learning, especially for training large-scale deep models.
On the Convergence of Memory-Based Distributed SGD
Using the transformation equation, we propose the convergence rate of stagewise M-DSGD which bridges the gap between theory and practice.
Collaborative Self-Attention for Recommender Systems
Recommender systems (RS), which have been an essential part in a wide range of applications, can be formulated as a matrix completion (MC) problem.
Deep Multi-Index Hashing for Person Re-Identification
In this paper, we propose a novel hashing method, called deep multi-index hashing (DMIH), to improve both efficiency and accuracy for ReID.
On the Evaluation Metric for Hashing
Bucket search, also called hash lookup, can achieve fast query speed with a sub-linear time cost based on the inverted index table constructed from hash codes.
Hashing based Answer Selection
HAS adopts a hashing strategy to learn a binary matrix representation for each answer, which can dramatically reduce the memory cost for storing the matrix representations of answers.
Gated Group Self-Attention for Answer Selection
Nevertheless, it is difficult for RNN based models to capture the information about long-range dependency among words in the sentences of questions and answers.
Quantized Epoch-SGD for Communication-Efficient Distributed Learning
Due to its efficiency and ease to implement, stochastic gradient descent (SGD) has been widely used in machine learning.
Proximal SCOPE for Distributed Sparse Learning
Based on the defined metric, we theoretically prove that pSCOPE is convergent with a linear convergence rate if the data partition is good enough.
Proximal SCOPE for Distributed Sparse Learning: Better Data Partition Implies Faster Convergence Rate
Based on the defined metric, we theoretically prove that pSCOPE is convergent with a linear convergence rate if the data partition is good enough.
Convolutional Geometric Matrix Completion
Recently, geometric deep learning on graphs (GDLG) is proposed to solve the GMC problem, showing better performance than existing GMC methods including traditional graph regularization based methods.
Feature-Distributed SVRG for High-Dimensional Linear Classification
Linear classification has been widely used in many high-dimensional applications like text classification.
Asymmetric Deep Supervised Hashing
However, most existing deep supervised hashing methods adopt a symmetric strategy to learn one deep hash function for both query points and database (retrieval) points.
Robust Frequent Directions with Application in Online Learning
We also apply RFD to online learning and propose an effective hyperparameter-free online Newton algorithm.
ParaGraphE: A Library for Parallel Knowledge Graph Embedding
Knowledge graph embedding aims at translating the knowledge graph into numerical representations by transforming the entities and relations into continuous low-dimensional vectors.
Full-Time Supervision based Bidirectional RNN for Factoid Question Answering
Hence, these existing models don't put supervision (loss or similarity calculation) at every time step, which will lose some useful information.
A Proximal Stochastic Quasi-Newton Algorithm
It incorporates the Hessian in the smooth part of the function and exploits multistage scheme to reduce the variance of the stochastic gradient.
SCOPE: Scalable Composite Optimization for Learning on Spark
Recently, many distributed stochastic optimization~(DSO) methods have been proposed to solve the large-scale composite optimization problems, which have shown better performance than traditional batch methods.
Feature Learning based Deep Supervised Hashing with Pairwise Labels
For another common application scenario with pairwise labels, there have not existed methods for simultaneous feature learning and hash-code learning.
A New Relaxation Approach to Normalized Hypergraph Cut
Normalized graph cut (NGC) has become a popular research topic due to its wide applications in a large variety of areas like machine learning and very large scale integration (VLSI) circuit design.
A Parallel algorithm for $\mathcal{X}$-Armed bandits
To deal with these large-scale data sets, we study a distributed setting of $\mathcal{X}$-armed bandits, where $m$ players collaborate to find the maximum of the unknown function.
Fast Asynchronous Parallel Stochastic Gradient Decent
Stochastic gradient descent~(SGD) and its variants have become more and more popular in machine learning due to their efficiency and effectiveness.
Scalable Stochastic Alternating Direction Method of Multipliers
There exists only one stochastic method, called SA-ADMM, which can achieve convergence rate $O(1/T)$ on general convex problems.
Distributed Power-law Graph Computing: Theoretical and Empirical Analysis
We theoretically prove that DBH can achieve lower communication cost than existing methods and can simultaneously guarantee good workload balance.
Supervised hashing with latent factor models
Very recently, supervised hashing methods, which try to preserve the semantic structure constructed from the semantic labels of the training points, have exhibited higher accuracy than unsupervised methods.
Isotropic Hashing
Most existing hashing methods adopt some projection functions to project the original data into several dimensions of real values, and then each of these projected dimensions is quantized into one bit (zero or one) by thresholding.
