论文整理集合 -- 吴恩达老师深度学习课程
2018-02-24 13:18
495 查看
吴恩达老师深度学习课程中所提到的论文整理集合!这些论文是深度学习的基本知识,阅读这些论文将更深入理解深度学习。
这些论文基本都可以免费下载到,如果无法免费下载,请留言!可以到coursera中看该视频课程。
下面的论文主要分为3部分:
1. 优化神经网络的各种方法
2. 卷积神经网络,包含各种model和物体检测的论文
3. RNN类型的神经网络
Srivastava, Nitish, et al. “Dropout: A simple way to prevent neural networks from overfitting.” The Journal of Machine Learning Research 15.1 (2014): 1929-1958.
https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf
RMSprop; optimization of gradient descent, it is an unpublished, adaptive learning rate method proposed by Geoff Hinton in Lecture 6e of his Coursera Class. RMSprop and Adadelta have both been developed independently around the same time stemming from the need to resolve Adagrad’s radically diminishing learning rates.
Tieleman, Tijmen, and Geoffrey Hinton. “Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude.” COURSERA: Neural networks for machine learning 4.2 (2012): 26-31.
Adam optimization algorithm; an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments.
https://arxiv.org/pdf/1412.6980.pdf
Batch normalization
Ioffe, Sergey, and Christian Szegedy. “Batch normalization: Accelerating deep network training by reducing internal covariate shift.” International conference on machine learning. 2015.
LeCun et al., 1998. Gradient-based learning applied to document recognition
- AlexNet; a kind of nueral network model;
Krizhevsky et al., 2012. ImageNet classification with deep convolutional neural networks
- VGG-16;
Simonyan & Zisserman 2015. Very deep convolutional networks for large-scale image recognition
ResNet(Residual Network);
He et al., 2015. Deep residual networks for image recognition
Network in Network (one by one convolution); filter size is (1 ,1), but filter number is more than one;
Lin et al., 2013, Network in network.
inception network; motivation for inception network;
Szegedy et al. 2014. Going deeper with convolutions
object recognition;
Sermanet et al., 2014, OverFeat: Integrated recognition, localization and detection using convolutional networks
YOLO (you only look once); real-time object detection;
Redmon et al,. 2015. You Only Look Once: Unified real-time object detection.
R-CNN; region proposal, classify proposed regions one at a time. output label + bounding box;
Girshik et al., 2013. Rich feature hierarchies for accurate object detection and semantic segmentation.
Fast R-CNN; Propose regions, use convolution implementation of sliding windows to classify all the proposed regions;
Girshik, 2015. Fast R-CNN.
Faster R-CNN; use convolutional network to propose regions;
Ren et.al, 2016. Faster R-CNN:Towards real-time object detection with region proposal networks.
Siamese network; Face recognition;
Taigman et.al., 2014. DeepFace closing the gap to human level performance
FaceNet;
Schreff et.al., 2015, FaceNet: A unified embedding for face recognition and clustering
Cho et al., 2014. On the properties of neural machine translation: Encoder-decoder approaches
gated recurrent unit;
Chung et al., 2014. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling.
LSTM (long short-term memory);
Hochreiter & Schmidhuber 1997. Long short-term memory
Visualizing word embeddings
van der Maaten and Hinton., 2008. Visualizing data using t-SNE
About word embedding
Mikolov et.al., 2013. Linguistic regularities in continuous space word representations
neural language model. to predict next word.
Bengio et.al., 2003, A neural probabilistic language model
Skip-gram model, about how to learn word-to-vector of word embedding in the neural network.
Mikolov et.al., 2013. Efficient estimation of word representations in vector space
Negative sampling; similar to skip-gram model but with much more efficient.
Mikolov et.al., 2013. Distributed representation of words and phrases and their compositionality.
GloVe (global vectors for word representation); Has some momentum in the NLP community. It is not used as much as the Word2Vec or the skip-gram models.
Pennington et.al., 2014. GloVe: Global vectors for word representation.
About the problem of bias in word embeddings.
Bolukbasi et.al., 2016. Man is to computer programmer as woman is to homemaker? Debiasing word embeddings
CTC (Connectionist temporal classification) cost for speech recognition
Graves et al., 2006. Connectionist Temporal Classification: Labeling unsegmented sequence data with recurrent neural networks
language tranlation; Sequence to sequence model
Sutskever et al., 2014. Sequence to sequence learning with neural networks
language tranlation; Sequence to sequence model
Cho et al., 2014. Learning phrase representations using RNN encoder-decoder for statistical machine translation
Image captioning
Mao et. al., 2014. Deep captioning with multimodal recurrent neural networks
Vinyals et.al., 2014. Show and tell: Neural image caption generator
Karpathy and Fei Fei, 2015. Deep visual-semantic alignments for generating image descriptions
Evaluating machine translation
Papineni et.al., 2002. A method for automatic evaluation of machine translation
Attention model
Bahdanau et.al., 2014. Neural machine translation by jointly learning to align and tranlate
Xu et.al., 2015. Show attention and tell: neural image caption generation with visual attention
注意力模型,第二张图是第一张图中“Attention” 的细分。
a<t,Tx>a<t,Tx>是a<Tx>a<Tx>的权值。其中a<t,Tx>a<t,Tx>越大,则对应的a<Tx>a<Tx>被注意的程度也就也大。
这些论文基本都可以免费下载到,如果无法免费下载,请留言!可以到coursera中看该视频课程。
下面的论文主要分为3部分:
1. 优化神经网络的各种方法
2. 卷积神经网络,包含各种model和物体检测的论文
3. RNN类型的神经网络
A collection of papers mentioned in the deep learning course of Andrew Ng
1. Neural Networks and Deep Learning
None2. Improving Deep Neural Networks Hyperparameter tuning, Regularization and Optimization
Dropout; regularization;Srivastava, Nitish, et al. “Dropout: A simple way to prevent neural networks from overfitting.” The Journal of Machine Learning Research 15.1 (2014): 1929-1958.
https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf
RMSprop; optimization of gradient descent, it is an unpublished, adaptive learning rate method proposed by Geoff Hinton in Lecture 6e of his Coursera Class. RMSprop and Adadelta have both been developed independently around the same time stemming from the need to resolve Adagrad’s radically diminishing learning rates.
Tieleman, Tijmen, and Geoffrey Hinton. “Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude.” COURSERA: Neural networks for machine learning 4.2 (2012): 26-31.
Adam optimization algorithm; an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments.
https://arxiv.org/pdf/1412.6980.pdf
Batch normalization
Ioffe, Sergey, and Christian Szegedy. “Batch normalization: Accelerating deep network training by reducing internal covariate shift.” International conference on machine learning. 2015.
3. Structuring Machine Learning Projects
None4. Convolutional Neural Networks
LeNet-5; a kind of nueral network model;LeCun et al., 1998. Gradient-based learning applied to document recognition
- AlexNet; a kind of nueral network model;
Krizhevsky et al., 2012. ImageNet classification with deep convolutional neural networks
- VGG-16;
Simonyan & Zisserman 2015. Very deep convolutional networks for large-scale image recognition
ResNet(Residual Network);
He et al., 2015. Deep residual networks for image recognition
Network in Network (one by one convolution); filter size is (1 ,1), but filter number is more than one;
Lin et al., 2013, Network in network.
inception network; motivation for inception network;
Szegedy et al. 2014. Going deeper with convolutions
object recognition;
Sermanet et al., 2014, OverFeat: Integrated recognition, localization and detection using convolutional networks
YOLO (you only look once); real-time object detection;
Redmon et al,. 2015. You Only Look Once: Unified real-time object detection.
R-CNN; region proposal, classify proposed regions one at a time. output label + bounding box;
Girshik et al., 2013. Rich feature hierarchies for accurate object detection and semantic segmentation.
Fast R-CNN; Propose regions, use convolution implementation of sliding windows to classify all the proposed regions;
Girshik, 2015. Fast R-CNN.
Faster R-CNN; use convolutional network to propose regions;
Ren et.al, 2016. Faster R-CNN:Towards real-time object detection with region proposal networks.
Siamese network; Face recognition;
Taigman et.al., 2014. DeepFace closing the gap to human level performance
FaceNet;
Schreff et.al., 2015, FaceNet: A unified embedding for face recognition and clustering
5. Sequence Models
gated recurrent unit;Cho et al., 2014. On the properties of neural machine translation: Encoder-decoder approaches
gated recurrent unit;
Chung et al., 2014. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling.
LSTM (long short-term memory);
Hochreiter & Schmidhuber 1997. Long short-term memory
Visualizing word embeddings
van der Maaten and Hinton., 2008. Visualizing data using t-SNE
About word embedding
Mikolov et.al., 2013. Linguistic regularities in continuous space word representations
neural language model. to predict next word.
Bengio et.al., 2003, A neural probabilistic language model
Skip-gram model, about how to learn word-to-vector of word embedding in the neural network.
Mikolov et.al., 2013. Efficient estimation of word representations in vector space
Negative sampling; similar to skip-gram model but with much more efficient.
Mikolov et.al., 2013. Distributed representation of words and phrases and their compositionality.
GloVe (global vectors for word representation); Has some momentum in the NLP community. It is not used as much as the Word2Vec or the skip-gram models.
Pennington et.al., 2014. GloVe: Global vectors for word representation.
About the problem of bias in word embeddings.
Bolukbasi et.al., 2016. Man is to computer programmer as woman is to homemaker? Debiasing word embeddings
CTC (Connectionist temporal classification) cost for speech recognition
Graves et al., 2006. Connectionist Temporal Classification: Labeling unsegmented sequence data with recurrent neural networks
language tranlation; Sequence to sequence model
Sutskever et al., 2014. Sequence to sequence learning with neural networks
language tranlation; Sequence to sequence model
Cho et al., 2014. Learning phrase representations using RNN encoder-decoder for statistical machine translation
Image captioning
Mao et. al., 2014. Deep captioning with multimodal recurrent neural networks
Vinyals et.al., 2014. Show and tell: Neural image caption generator
Karpathy and Fei Fei, 2015. Deep visual-semantic alignments for generating image descriptions
Evaluating machine translation
Papineni et.al., 2002. A method for automatic evaluation of machine translation
Attention model
Bahdanau et.al., 2014. Neural machine translation by jointly learning to align and tranlate
Xu et.al., 2015. Show attention and tell: neural image caption generation with visual attention
注意力模型,第二张图是第一张图中“Attention” 的细分。
a<t,Tx>a<t,Tx>是a<Tx>a<Tx>的权值。其中a<t,Tx>a<t,Tx>越大,则对应的a<Tx>a<Tx>被注意的程度也就也大。
相关文章推荐
- 吴恩达老师在course上开设了一门深度学习的课程的作业
- Operations on word vectors-v2 吴恩达老师深度学习课程第五课第二周编程作业1
- 吴恩达深度学习课程心得
- 吴恩达Coursera深度学习课程 DeepLearning.ai 编程作业——Keras tutorial - the Happy House (4.2)
- 吴恩达Coursera深度学习课程 DeepLearning.ai 编程作业——Autonomous driving - Car detection(4.3)
- 深度增强学习方向论文整理
- 吴恩达 深度学习 Class1 课程总结及编程实践
- 干货丨吴恩达深度学习课程的思维导图总结
- 深度学习论文整理
- 吴恩达课程深度学习错题集
- DLRS(深度学习应用于推荐系统论文汇总--2017年8月整理)
- 吴恩达Coursera深度学习课程 DeepLearning.ai 提炼笔记(1-2)-- 神经网络基础
- 吴恩达Coursera深度学习课程 DeepLearning.ai 编程作业——Optimization Methods(2-2)
- 吴恩达深度学习课程deeplearning.ai课程作业:Class 2 Week 3 TensorFlow Tutorial
- 吴恩达深度学习课程笔记之卷积神经网络基本操作详解
- 学完吴恩达全部深度学习课程,这有一份课程解读
- 吴恩达Coursera深度学习课程 DeepLearning.ai 提炼笔记(2-2)-- 优化算法
- 吴恩达Coursera深度学习课程 DeepLearning.ai 提炼笔记(4-1)-- 卷积神经网络基础
- 吴恩达深度学习课程第二部分笔记要点
- 吴恩达深度学习课程笔记-1.