deep-neural-networks noisy-data label-noise noisy-labels unreliable-labels robust-learning

Awesome-Learning-with-Label-Noise:精选的带有噪音标签的学习资源列表

标签： deep-neural-networks noisy-data label-noise noisy-labels unreliable-labels robust-learning

A curated list of resources for Learning with Noisy Labels 论文和代码 2008年NIPS-谁的选票应该更多：未知专业贴标商对标签的最佳整合。 2009-ICML-在多位专家的监督下学习：每个人都撒谎时值得信任的人。 ...

Building your Deep Neural Network - Step by Step v5 作业 - Neural Networks and Deep Learning

标签：深度学习 Neural Networks

复习用Building your Deep Neural Network: Step by StepWelcome to your week 4 assignment (part 1 of 2)! You have previously trained a 2-layer Neural Network (with a single hidden layer). This w

19-Self-supervised-Visual-Feature-Learning-with-Deep-Neural-Networks-A-Survey

who Longlong Jing and Yingli Tian ∗ , Fellow, IEEE 2019- what 为了避免收集和注释大规模数据集的大量成本，作为无监督学习方法的子集，提出了...1. Pseudo label: 伪标签是基于pretext tasks的数据属性自动...

Coursera-Deep Learning Specialization 课程之（一）：Neural Networks and Deep Learning-weak4编程作业

标签： deep-learning

Building your Deep Neural Network: Step by Step1 - Packagesimport numpy as np import h5py import matplotlib.pyplot as plt from testCases_v3 import * from dnn_utils_v2 import sigmoid, sigmoid_backward,

Paper reading (四十一)：Deep Learning in Label-free Cell Classification

标签： DNN AUC label-free cell classification

论文题目：Deep Learning in Label-free Cell Classification scholar 引用：190 页数：16 发表时间：2016.03 发表刊物：nature scientific reports 作者：Claire Lifan Chen, Ata Mahjoubfar,..., Bahram ...

Coursera-Deep Learning Specialization 课程之（一）：Neural Networks and Deep Learning-weak2编程作业

标签： deep-learning python

一 Python Basics with numpy (optional)学习目标： ①使用logistic regression ②学习如何最小化代价函数cost function ③理解通过代价函数的导数来更新参数**1 - Building basic functions with numpy1.1 - ...

《每日论文》ImageNet Classification with Deep Convolutional Neural Networks

这篇论文是剖析 CNN 领域的经典之作，也是入门 CNN 的必读论文。作者训练了一个面向数量为 1.2 百万的高分辨率的图像数据集 ImageNet，图像的...ImageNet Classification with Deep Convolutional Neural Networks .

【Noise-Label】《Learning from Noisy Labels with Deep Neural Networks》

标签： bottom-up top-down re-weight

文章目录1 Background and Motivation2 Advantages3 Innovations4 Method4.1 Bottom-up Noise Model4.2 Estimating Noise Distribution Using Clean Data4.3 Learning Noise Distribution From Noisy Data...

【论文整理】小样本学习Few-shot learning论文整理收藏（最全，持续更新）

标签：小样本学习综述 few-shot 小样本

1.Generalizing from a Few Examples: A Survey on Few-Shot Learning 2.Generalizing from a few examples: A survey on few-shot learning, CSUR, 2020. 3.Rethinkingfew-shotimage classification: a good ...

IDN:AAAI 2021

标签： deep-neural-networks noisy-data label-noise noisy-labels robust-learning instance-dependent-noise Python

title={Beyond Class-Conditional Assumption: A Primary Attempt to Combat Instance-Dependent Label Noise.}, author={Chen, Pengfei and Ye, Junjie and Chen, Guangyong and Zhao, Jingwei and Heng, Pheng-...

RobustnessAccuracy:AAAI 2021

标签： deep-neural-networks validation noisy-data label-noise noisy-labels early-stopping robust-learning Python

title={Robustness of Accuracy Metric and its Inspirations in Learning with Noisy Labels.}, author={Chen, Pengfei and Ye, Junjie and Chen, Guangyong and Zhao, Jingwei and Heng, Pheng-Ann}, journal={...

【multi-label】Learning a Deep ConvNet for Multi-label Classification with Partial Labels

标签： Multi-lables partial lable noisy lable

来源：CVPR2019 ... 本文目的：为了减少图片中多标签标注的成本，提出了一种训练模型的方式：训练模型的样本使用标签没有标注完整的图片；作者的贡献： 1）比较了多标签数据集的标注方法，作者的方法证明了使用部分...

遥感图像-Deep Feature Alignment Neural Networks for Domain Adaptation of Hyperspectral Data高光谱...

本文中,我们建议使用深特性结合神经网络进行域的适应,在标记辅助数据源的数据可以用来提高分类的性能在有限领域,否则带安全标签的数据是可用的。该模型首先利用深度卷积递归神经网络提取源和目标域的鉴别特征，然后...

LLM高效参数微调方法：从Prefix Tuning、Prompt Tuning、P-Tuning V1/V2到LoRA、QLoRA(含对模型量化的解释)

标签：大模型微调方法 Prefix Tuning P-Tuning V1/V2

常规部分的正向传播由transformers所定义，而LoRA部分的正向传播则由LinearLayer_LoRA(nn.Module)的forward()所定义，即“LoRA层的两条分支结果进行加和”，如下图所示『一般用随机高斯分布初始化，当然实际代码实现...

【CV】高被引行人重识别（Person Re-ID）综述论文

标签：计算机视觉行人重识别 Re-ID

论文年份：TPAMI 2021，论文被引：448（2022/05/17）

Multi-task Learning(Review)多任务学习概述

标签： Multi-task Learning(Review)多任务

背景：只专注于单个模型可能会忽略一些相关任务中可能提升目标任务的潜在信息，通过进行一定程度的...广义的讲，只要loss有多个就算MTL，一些别名（joint learning，learning to learn，learning with auxil...

2022-03-04-Label-Noise-Learning

标签：机器学习人工智能深度学习

layout comments title categories tags description post true Label Noise Learning blog 自然语言处理 Label Noise Learning L...

mentornet：MentorNet的代码：学习非常深的神经网络的数据驱动课程

标签： google deep-learning label noisy-data noisy GooglePython

如果您发现此代码对您的研究有用，请引用@inproceedings{jiang2018mentornet, title={MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels}, author={Jiang, Lu and ...

喜p：用于多标签文本分类的深度神经网络框架

标签： nlp machine-learning deep-learning neural-network word2vec prediction classification multi-label-classification NaturallanguageprocessingPython

喜p是用于多标签文本分类的深度学习工具。它在训练语料库上学习将标签分配给任意文本，... 可以在data / hep-categories目录下找到示例语料库。 .lab查找包含文本以进行预测的.txt文件，并在单独的行中分配标签的.lab

Paper reading (三十)：A review on machine learning principles for multi-view biological data ...

标签： data integration multi-omics data machine learning

论文题目：A review on machine learning principles for multi-view biological data integration scholar 引用：69 页数：16 发表时间：2016.12 发表刊物：Briefings in Bioinformatics 作者：Yifeng Li, ...

Pseudo-Label : The Simple and Efficient Semi-Supervised Learning Method for Deep Neural Networks

标签：机器学习算法深度学习

提出了一种简单有效的深神经网络半监督学习方法。基本上，所提出的网络是以有监督的方式训练，同时有标记和无标记的数据。对于未标记的数据，伪标签，只是选取具有最大预测概率的类，就好像它们是真的...Pseudo-Label.

GNN algorithms(3): Tri-party Deep Network Representation

标签： Tri-DNR DeepWalk

conceptrepresent each node 表示节点 in a vector format rather than others。Aim：to learn.learn optimal node representation. 节点信息在本论文中不是一个单独的概念，而是概括性的概念。

【论文精读】MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted ...

MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels 原文地址：MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels ...

“Physics-based Deep Learning”学习笔记(1)

标签：深度学习神经网络机器学习

“Physics-based Deep Learning”学习笔记1 可微分物理学(Differentiable physics(DP)) 基于物理的深度学习，其本质是应用某个特定领域内的模型方程，将其离散化之后的方程整合进深度学习的训练过程。正如其名称中所...

文献解读-Physics Informed Deep Learning(PINN)

标签：深度学习机器学习偏微分方程

Physics Informed Deep Learning Data-driven solutions and discovery of Nonlinear Partial Differential Equations View on GitHub Authors Maziar Raissi, Paris Perdikaris, and George Em Karniadakis ...

吴恩达 Deep Learning作业：(课程1-第4周) - Deep+Neural+Network+-+Application+v8

标签：吴恩达 Deep Learning

Deep Neural Network for Image Classification: Application¶ When you finish this, you will have finished the last programming assignment of Week 4, and also the last programming assignment of this ...

DeepLearing学习笔记-Building your Deep Neural Network: Step by Step(第四周作业)

标签：神经网络 deep-learning

1-背景：此前，我们已经介绍过单隐藏层的神经网络模型，本文要介绍的是多隐藏层的神经网络模型。采用非线性的如RELU激活函数符号说明：上标 [l][l] 表示层号，lthl^{th} 例如: a[L]a^{[L]} 是第 LthL^{th} 层的...

【论文阅读】Pseudo-Label : The Simple and Eﬃcient Semi-Supervised Learning Method for Deep Neural ...

标签：深度学习半监督学习 DAE

【论文阅读】Pseudo-Label : The Simple and Eﬃcient Semi-Supervised Learning Method for Deep Neural Networks

【TIP 2020】Iterative Local-Global Collaboration Learning Towards One-Shot Video Person Re-ID

标签：深度学习 python 行人重识别

引言提出了local-global 协作学习方法，用来进行标签估计在损失中引入了变量信息瓶颈作为正则项。使得特征提取器能过滤掉无关因素采用了和eug一样的迭代模式。文章目录引言内容概要工作概述成果概述方法详解...

元学习之《Siamese Neural Networks for One-shot Image Recognition》代码解读

标签： meta-learning 元学习少样本学习

《Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks》论文翻译笔记元学习方向 optimization based meta learning 之 MAML论文详细解读 MAML 源代码解释说明 (一) MAML 源代码解释说明 (二) 元...