About this course: Probabilistic graphical models (PGMs) are a rich framework for encoding probability distributions over complex domains: joint (multivariate) distributions over large numbers of random variables that interact with each other. These representations sit at the intersection of statistics and computer science, relying on concepts from probability theory, graph algorithms, machine learning, and more. They are the basis for the state-of-the-art methods in a wide variety of applications, such as medical diagnosis, image understanding, speech recognition, natural language processing, and many, many more. They are also a foundational tool in formulating many machine learning problems. This course is the first in a sequence of three. It describes the two basic PGM representations: Bayesian Networks, which rely on a directed graph; and Markov networks, which use an undirected graph. The course discusses both the theoretical properties of these representations as well as their use in practice. The (highly recommended) honors track contains several hands-on assignments on how to represent some real-world problems. The course also presents some important extensions beyond the basic PGM representation, which allow more complex models to be encoded compactly.
Created by: Stanford University
Taught by: Daphne Koller, Professor
School of Engineering
| Basic Info | Course 1 of 3 in the Probabilistic Graphical Models Specialization |
| Level | Advanced |
| Language | English |
| How To Pass | Pass all graded assignments to complete the course. |
| User Ratings |
Syllabus
WEEK 1
Introduction and Overview
This module provides an overall introduction to probabilistic graphical models, and defines a few of the key concepts that will be used later in the course.
4 videos
- Video: Welcome!
- Video: Overview and Motivation
- Video: Distributions
- Video: Factors
Graded: Basic Definitions
Bayesian Network (Directed Models)
In this module, we define the Bayesian network representation and its semantics. We also analyze the relationship between the graph structure and the independence properties of a distribution represented over that graph. Finally, we give some practical tips on how to model a real-world situation as a Bayesian network.
15 videos, 6 readings
- Video: Semantics & Factorization
- Video: Reasoning Patterns
- Video: Flow of Probabilistic Influence
- Video: Conditional Independence
- Video: Independencies in Bayesian Networks
- Video: Naive Bayes
- Video: Application - Medical Diagnosis
- Video: Knowledge Engineering Example - SAMIAM
- Reading: Setting Up Your Programming Assignment Environment
- Reading: Installing Octave/MATLAB on Windows
- Reading: Installing Octave/MATLAB on Mac OS X (10.10 Yosemite and 10.9 Mavericks)
- Reading: Installing Octave/MATLAB on Mac OS X (10.8 Mountain Lion and Earlier)
- Reading: Installing Octave/MATLAB on GNU/Linux
- Reading: More Octave/MATLAB resources
- Video: Basic Operations
- Video: Moving Data Around
- Video: Computing On Data
- Video: Plotting Data
- Video: Control Statements: for, while, if statements
- Video: Vectorization
- Video: Working on and Submitting Programming Exercises
Graded: Bayesian Network Fundamentals
Graded: Bayesian Network Independencies
Graded: Octave/Matlab installation
Graded: Simple BN Knowledge Engineering
WEEK 2
Template Models for Bayesian Networks
In many cases, we need to model distributions that have a recurring structure. In this module, we describe representations for two such situations. One is temporal scenarios, where we want to model a probabilistic structure that holds constant over time; here, we use Hidden Markov Models, or, more generally, Dynamic Bayesian Networks. The other is aimed at scenarios that involve multiple similar entities, each of whose properties is governed by a similar model; here, we use Plate Models.
4 videos
- Video: Overview of Template Models
- Video: Temporal Models - DBNs
- Video: Temporal Models - HMMs
- Video: Plate Models
Graded: Template Models
Structured CPDs for Bayesian Networks
A table-based representation of a CPD in a Bayesian network has a size that grows exponentially in the number of parents. There are a variety of other form of CPD that exploit some type of structure in the dependency model to allow for a much more compact representation. Here we describe a number of the ones most commonly used in practice.
4 videos
- Video: Overview: Structured CPDs
- Video: Tree-Structured CPDs
- Video: Independence of Causal Influence
- Video: Continuous Variables
Graded: Structured CPDs
Graded: BNs for Genetic Inheritance
Graded: BNs for Genetic Inheritance PA Quiz
WEEK 3
Markov Networks (Undirected Models)
In this module, we describe Markov networks (also called Markov random fields): probabilistic graphical models based on an undirected graph representation. We discuss the representation of these models and their semantics. We also analyze the independence properties of distributions encoded by these graphs, and their relationship to the graph structure. We compare these independencies to those encoded by a Bayesian network, giving us some insight on which type of model is more suitable for which scenarios.
7 videos
- Video: Pairwise Markov Networks
- Video: General Gibbs Distribution
- Video: Conditional Random Fields
- Video: Independencies in Markov Networks
- Video: I-maps and perfect maps
- Video: Log-Linear Models
- Video: Shared Features in Log-Linear Models
Graded: Markov Networks
Graded: Independencies Revisited
Graded: Markov Networks for OCR
WEEK 4
Decision Making
In this module, we discuss the task of decision making under uncertainty. We describe the framework of decision theory, including some aspects of utility functions. We then talk about how decision making scenarios can be encoded as a graphical model called an Influence Diagram, and how such models provide insight both into decision making and the value of information gathering.
3 videos
- Video: Maximum Expected Utility
- Video: Utility Functions
- Video: Value of Perfect Information
Graded: Decision Theory
Graded: Decision Making
Graded: Decision Making PA Quiz
WEEK 5
Knowledge Engineering & Summary
This module provides an overview of graphical model representations and some of the real-world considerations when modeling a scenario as a graphical model. It also includes the course final exam.
1 video
- Video: Knowledge Engineering
Graded: Representation Final Exam
FAQs
How It Works
Coursework
Each course is like an interactive textbook, featuring pre-recorded videos, quizzes and projects.
Help from Your Peers
Connect with thousands of other learners and debate ideas, discuss course material, and get help mastering concepts.
Certificates
Earn official recognition for your work, and share your success with friends, colleagues, and employers.
Creators
Stanford University
The Leland Stanford Junior University, commonly referred to as Stanford University or Stanford, is an American private research university located in Stanford, California on an 8,180-acre (3,310 ha) campus near Palo Alto, California, United States.
Pricing
| Audit | Purchase Course | |
|---|---|---|
| Access to course materials | Available | Available |
| Access to graded materials | Not available | Available |
| Receive a final grade | Not available | Available |
| Earn a shareable Course Certificate | Not available | Available |
Ratings and Reviews
I guess this is probably the most challenging one in the Coursera. Really Hard but really rewarding course!
Awsome course for Information/Knowledge Engineering. Although not necessary to finish all the honor assignments, it is highly recommended to implement them. Not only for comprehension, but also practice. You can actually apply them on your career or research.
Awesome material. Could not get this experience by learning the subject ourselves using a textbook.
Coursera provides universal access to the world’s best education, partnering with top universities and organizations to offer courses online.
© 2017 Coursera Inc. All rights reserved.
AE
Covers some material a little too quickly, but overall a good and entertaining course.