course1:Neural Networks and Deep Learning
c1_week1: Introduction to deep learning
Be able to explain the major trends driving the rise of deep learning, and understand where and how it is applied today.
学习目标
- Understand the major trends driving the rise of deep learning.
- Be able to explain how deep learning is applied to supervised learning.
- Understand what are the major categories of models (such as CNNs and RNNs), and when they should be applied.
- Be able to recognize the basics of when deep learning will (or will not) work well.
Welcome to the Deep Learning Specialization
课程视频Welcome
Introduction to Deep Learning
课程视频What is a neural network?
课程视频Supervised Learning with Neural Networks
课程视频Why is Deep Learning taking off?
课程视频About this Course
阅读材料Frequently Asked Questions
课程视频Course Resources
阅读材料How to use Discussion Forums
c1_week2:Neural Networks Basics
Learn to set up a machine learning problem with a neural network mindset. Learn to use vectorization to speed up your models.
学习目标
- Build a logistic regression model, structured as a shallow neural network
- Implement the main steps of an ML algorithm, including making predictions, derivative computation, and gradient descent.
- Implement computationally efficient, highly vectorized, versions of models.
- Understand how to compute derivatives for logistic regression, using a backpropagation mindset.
- Become familiar with Python and Numpy
- Work with iPython Notebooks
- Be able to implement vectorization across multiple training examples
Logistic Regression as a Neural Network
课程视频Binary Classification
课程视频Logistic Regression
课程视频Logistic Regression Cost Function
课程视频Gradient Descent
课程视频Derivatives
课程视频More Derivative Examples
课程视频Computation graph
课程视频Derivatives with a Computation Graph
课程视频Logistic Regression Gradient Descent
课程视频Gradient Descent on m Examples
Python and Vectorization
课程视频Vectorization
课程视频More Vectorization Examples
课程视频Vectorizing Logistic Regression
课程视频Vectorizing Logistic Regression's Gradient Output
课程视频Broadcasting in Python
课程视频A note on python/numpy vectors
课程视频Quick tour of Jupyter/iPython Notebooks
课程视频Explanation of logistic regression cost function (optional)
Programming Assignments
编程作业:Python Basics with numpy (optional)
编程作业: Logistic Regression with a Neural Network mindset
c1_week3: Shallow neural networks
Learn to build a neural network with one hidden layer, using forward propagation and backpropagation.
学习目标
- Understand hidden units and hidden layers
- Be able to apply a variety of activation functions in a neural network.
- Build your first forward and backward propagation with a hidden layer
- Apply random initialization to your neural network
- Become fluent with Deep Learning notations and Neural Network Representations
- Build and train a neural network with one hidden layer.
Shallow Neural Network
课程视频Neural Networks Overview
课程视频Neural Network Representation
课程视频Computing a Neural Network's Output
课程视频Vectorizing across multiple examples
课程视频Explanation for Vectorized Implementation
课程视频Activation functions
课程视频Why do you need non-linear activation functions?
课程视频Derivatives of activation functions
课程视频Gradient descent for Neural Networks
课程视频Backpropagation intuition (optional)
课程视频Random Initialization
Programming Assignment
编程作业: Planar data classification with a hidden layer
c1_week4: Deep Neural Networks
Understand the key computations underlying deep learning, use them to build and train deep neural networks, and apply it to computer vision.
学习目标
- See deep neural networks as successive blocks put one after each other
- Build and train a deep L-layer Neural Network
- Analyze matrix and vector dimensions to check neural network implementations.
- Understand how to use a cache to pass information from forward propagation to back propagation.
- Understand the role of hyperparameters in deep learning
Deep Neural Network
课程视频Deep L-layer neural network
课程视频Forward Propagation in a Deep Network
课程视频Getting your matrix dimensions right
课程视频Why deep representations?
课程视频Building blocks of deep neural networks
课程视频Forward and Backward Propagation
课程视频Parameters vs Hyperparameters
课程视频What does this have to do with the brain?
Programming Assignments
编程作业: Building your deep neural network: Step by Step
编程作业: Deep Neural Network Application
course2: Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization
c2_week1: Practical aspects of Deep Learning
学习目标
- Recall that different types of initializations lead to different results
- Recognize the importance of initialization in complex neural networks.
- Recognize the difference between train/dev/test sets
- Diagnose the bias and variance issues in your model
- Learn when and how to use regularization methods such as dropout or L2 regularization.
- Understand experimental issues in deep learning such as Vanishing or Exploding gradients and learn how to deal with them
- Use gradient checking to verify the correctness of your backpropagation implementation
Setting up your Machine Learning Application
课程视频Train / Dev / Test sets
课程视频Bias / Variance
课程视频Basic Recipe for Machine Learning
Regularizing your neural network
课程视频Regularization
课程视频Why regularization reduces overfitting?
课程视频Dropout Regularization
课程视频Understanding Dropout
课程视频Other regularization methods
Setting up your optimization problem
课程视频Normalizing inputs
课程视频Vanishing / Exploding gradients
课程视频Weight Initialization for Deep Networks
课程视频Numerical approximation of gradients
课程视频Gradient checking
课程视频Gradient Checking Implementation Notes
Programming assignments
编程作业: Initialization
编程作业: Regularization
编程作业: Gradient Checking
c2_week2: Optimization algorithms
学习目标
- Remember different optimization methods such as (Stochastic) Gradient Descent, Momentum, RMSProp and Adam
- Use random minibatches to accelerate the convergence and improve the optimization
- Know the benefits of learning rate decay and apply it to your optimization
Optimization algorithms
课程视频Mini-batch gradient descent
课程视频Understanding mini-batch gradient descent
课程视频Exponentially weighted averages
课程视频Understanding exponentially weighted averages
课程视频Bias correction in exponentially weighted averages
课程视频Gradient descent with momentum
课程视频RMSprop
课程视频Adam optimization algorithm
课程视频Learning rate decay
课程视频The problem of local optima
Programming assignment
编程作业: Optimization
c2_week3: Hyperparameter tuning, Batch Normalization and Programming Frameworks
学习目标
- Master the process of hyperparameter tuning
Hyperparameter tuning
课程视频Tuning process
课程视频Using an appropriate scale to pick hyperparameters
课程视频Hyperparameters tuning in practice: Pandas vs. Caviar
Batch Normalization
课程视频Normalizing activations in a network
课程视频Fitting Batch Norm into a neural network
课程视频Why does Batch Norm work?
课程视频Batch Norm at test time
Multi-class classification
课程视频Softmax Regression
课程视频Training a softmax classifier
Introduction to programming frameworks
课程视频Deep learning frameworks
课程视频TensorFlow
Programming assignment
编程作业: Tensorflow
course3: Structuring Machine Learning Projects
c3_week1: ML Strategy (1)
学习目标
- Understand why Machine Learning strategy is important
- Apply satisficing and optimizing metrics to set up your goal for ML projects
- Choose a correct train/dev/test split of your dataset
- Understand how to define human-level performance
- Use human-level perform to define your key priorities in ML projects
- Take the correct ML Strategic decision based on observations of performances and dataset
Introduction to ML Strategy
课程视频Why ML Strategy
课程视频Orthogonalization
Setting up your goal
课程视频Single number evaluation metric
课程视频Satisficing and Optimizing metric
课程视频Train/dev/test distributions
课程视频Size of the dev and test sets
课程视频When to change dev/test sets and metrics
Comparing to human-level performance
课程视频Why human-level performance?
课程视频Avoidable bias
课程视频Understanding human-level performance
课程视频Surpassing human-level performance
课程视频Improving your model performance
Machine Learning flight simulator
阅读材料Machine Learning flight simulator
测验: Bird recognition in the city of Peacetopia (case study)
c3_week2: ML Strategy (2)
学习目标
- Understand what multi-task learning and transfer learning are
- Recognize bias, variance and data-mismatch by looking at the performances of your algorithm on train/dev/test sets
Error Analysis
课程视频Carrying out error analysis
课程视频Cleaning up incorrectly labeled data
课程视频Build your first system quickly, then iterate
Mismatched training and dev/test set
课程视频Training and testing on different distributions
课程视频Bias and Variance with mismatched data distributions
课程视频Addressing data mismatch
Learning from multiple tasks
课程视频Transfer learning
课程视频Multi-task learning
End-to-end deep learning
课程视频What is end-to-end deep learning?
课程视频Whether to use end-to-end deep learning
Machine Learning flight simulator
测验: Autonomous driving (case study)
course4: Convolutional Neural Networks
c4_week1: Foundations of Convolutional Neural Networks
Learn to implement the foundational layers of CNNs (pooling, convolutions) and to stack them properly in a deep network to solve multi-class image classification problems.
学习目标
- Understand the convolution operation
- Understand the pooling operation
- Remember the vocabulary used in convolutional neural network (padding, stride, filter, ...)
- Build a convolutional neural network for image multi-class classification
Convolutional Neural Networks
课程视频Computer Vision
课程视频Edge Detection Example
课程视频More Edge Detection
课程视频Padding
课程视频Strided Convolutions
课程视频Convolutions Over Volume
课程视频One Layer of a Convolutional Network
课程视频Simple Convolutional Network Example
课程视频Pooling Layers
课程视频CNN Example
课程视频Why Convolutions?
Programming assignments
编程作业: Convolutional Model: step by step
编程作业: Convolutional model: application
c4_week2: Deep convolutional models: case studies
Learn about the practical tricks and methods used in deep CNNs straight from the research papers.
学习目标
- Understand multiple foundational papers of convolutional neural networks
- Analyze the dimensionality reduction of a volume in a very deep network
- Understand and Implement a Residual network
- Build a deep neural network using Keras
- Implement a skip-connection in your network
- Clone a repository from github and use transfer learning
Case studies
课程视频Why look at case studies?
课程视频Classic Networks
课程视频ResNets
课程视频Why ResNets Work
课程视频Networks in Networks and 1x1 Convolutions
课程视频Inception Network Motivation
课程视频Inception Network
Practical advices for using ConvNets
课程视频Using Open-Source Implementation
课程视频Transfer Learning
课程视频Data Augmentation
课程视频State of Computer Vision
Programming assignments
编程作业: Keras Tutorial - The Happy House (not graded)
编程作业: Residual Networks
c4_week3: Object detection
Learn how to apply your knowledge of CNNs to one of the toughest but hottest field of computer vision: Object detection.
学习目标
- Understand the challenges of Object Localization, Object Detection and Landmark Finding
- Understand and implement non-max suppression
- Understand and implement intersection over union
- Understand how we label a dataset for an object detection application
- Remember the vocabulary of object detection (landmark, anchor, bounding box, grid, ...)
Detection algorithms
课程视频Object Localization
课程视频Landmark Detection
课程视频Object Detection
课程视频Convolutional Implementation of Sliding Windows
课程视频Bounding Box Predictions
课程视频Intersection Over Union
课程视频Non-max Suppression
课程视频Anchor Boxes
课程视频YOLO Algorithm
课程视频(Optional) Region Proposals
Programming assignments
编程作业: Car detection with YOLOv2
c4_week4: Special applications: Face recognition & Neural style transfer
Discover how CNNs can be applied to multiple fields, including art generation and face recognition. Implement your own algorithm to generate art and recognize faces!
Face Recognition
课程视频What is face recognition?
课程视频One Shot Learning
课程视频Siamese Network
课程视频Triplet Loss
课程视频Face Verification and Binary Classification
Neural Style Transfer
课程视频What is neural style transfer?
课程视频What are deep ConvNets learning?
课程视频Cost Function
课程视频Content Cost Function
课程视频Style Cost Function
课程视频1D and 3D Generalizations
Programming assignments
编程作业: Art generation with Neural Style Transfer
编程作业: Face Recognition for the Happy House
course5: Sequence Models
c5_week1: Recurrent Neural Networks
Learn about recurrent neural networks. This type of model has been proven to perform extremely well on temporal data. It has several variants including LSTMs, GRUs and Bidirectional RNNs, which you are going to learn about in this section.
Recurrent Neural Networks
课程视频Why sequence models
课程视频Notation
课程视频Recurrent Neural Network Model
课程视频Backpropagation through time
课程视频Different types of RNNs
课程视频Language model and sequence generation
课程视频Sampling novel sequences
课程视频Vanishing gradients with RNNs
课程视频Gated Recurrent Unit (GRU)
课程视频Long Short Term Memory (LSTM)
课程视频Bidirectional RNN
课程视频Deep RNNs
Programming assignments
编程作业: Building a recurrent neural network - step by step
编程作业: Dinosaur Island - Character-Level Language Modeling
编程作业: Jazz improvisation with LSTM
c5_week2: Natural Language Processing & Word Embeddings
Natural language processing with deep learning is an important combination. Using word vector representations and embedding layers you can train recurrent neural networks with outstanding performances in a wide variety of industries. Examples of applications are sentiment analysis, named entity recognition and machine translation.
Introduction to Word Embeddings
课程视频Word Representation
课程视频Using word embeddings
课程视频Properties of word embeddings
课程视频Embedding matrix
Learning Word Embeddings: Word2vec & GloVe
课程视频Learning word embeddings
课程视频Word2Vec
课程视频Negative Sampling
课程视频GloVe word vectors
Applications using Word Embeddings
课程视频Sentiment Classification
课程视频Debiasing word embeddings
Programming assignments
编程作业: Operations on word vectors - Debiasing
编程作业: Emojify
c5_week3: Sequence models & Attention mechanism
Sequence models can be augmented using an attention mechanism. This algorithm will help your model understand where it should focus its attention given a sequence of inputs. This week, you will also learn about speech recognition and how to deal with audio data.