deep learnng summary

Deep Learning Tutorials

Deep Learning is a new area of Machine Learning research, which has been introduced with the objective of moving Machine Learning closer to one of its original goals: Artificial Intelligence. See these course notes for a brief introduction to Machine Learning for AI and an introduction to Deep Learning algorithms.

Deep Learning is about learning multiple levels of representation and abstraction that help to make sense of data such as images, sound, and text. For more about deep learning algorithms, see for example:

• The monograph or review paper Learning Deep Architectures for AI (Foundations & Trends in Machine Learning, 2009).
• The ICML 2009 Workshop on Learning Feature Hierarchies webpage has a list of references.
• The LISA public wiki has a reading list and a bibliography.
• Geoff Hinton has readings from 2009’s NIPS tutorial.

The tutorials presented here will introduce you to some of the most important deep learning algorithms and will also show you how to run them using Theano. Theano is a python library that makes writing deep learning models easy, and gives the option of training them on a GPU.

The algorithm tutorials have some prerequisites. You should know some python, and be familiar with numpy. Since this tutorial is about using Theano, you should read over the Theano basic tutorial first. Once you’ve done that, read through our Getting Startedchapter – it introduces the notation, and downloadable datasets used in the algorithm tutorials, and the way we do optimization by stochastic gradient descent.

The code is available on the Deep Learning Tutorial repositories.

The purely supervised learning algorithms are meant to be read in order:

1. Logistic Regression - using Theano for something simple
2. Multilayer perceptron - introduction to layers
3. Deep Convolutional Network - a simplified version of LeNet5

The unsupervised and semi-supervised learning algorithms can be read in any order (the auto-encoders can be read independently of the RBM/DBN thread):

• Auto Encoders, Denoising Autoencoders - description of autoencoders
• Stacked Denoising Auto-Encoders - easy steps into unsupervised pre-training for deep nets
• Restricted Boltzmann Machines - single layer generative RBM model
• Deep Belief Networks - unsupervised generative pre-training of stacked RBMs followed by supervised fine-tuning

Building towards including the mcRBM model, we have a new tutorial on sampling from energy models:

• HMC Sampling - hybrid (aka Hamiltonian) Monte-Carlo sampling with scan()

Building towards including the Contractive auto-encoders tutorial, we have the code for now:

• Contractive auto-encoders code - There is some basic doc in the code.

Recurrent neural networks with word embeddings and context window:

• Semantic Parsing of Speech using Recurrent Net

LSTM network for sentiment analysis:

• LSTM network

Energy-based recurrent neural network (RNN-RBM):

• Modeling and generating sequences of polyphonic music

Note that the tutorials here are all compatible with Python 2 and 3, with the exception of Modeling and generating sequences of polyphonic music with the RNN-RBM which is only available for Python 2.

线性代数、概率和信息论

欠拟合、过拟合、正则化

最大似然估计和贝叶斯统计

随机梯度下降

监督学习和无监督学习

深度前馈网络、代价函数和反向传播

正则化、稀疏编码和dropout

自适应学习算法

卷积神经网络

循环神经网络

递归神经网络

深度神经网络和深度堆叠网络

LSTM长短时记忆

主成分分析

正则自动编码器

表征学习

蒙特卡洛

受限波兹曼机

深度置信网络

softmax回归、决策树和聚类算法

KNN和SVM

生成对抗网络和有向生成网络

机器视觉和图像识别

自然语言处理

语音识别和机器翻译

有限马尔科夫

动态规划

梯度策略算法

增强学习（Q-learning）