Unsupervised Machine Learning: Hidden Markov Models in Python

HMMs for stock price analysis, language modeling, web analytics, biology, and PageRank.

Register for this Course

$29.99 $199.99 USD 85% OFF!

Login or signup to register for this course

Have a coupon? Click here.

Course Data

Lectures: 75
Length: 10h 39m
Skill Level: All Levels
Languages: English
Includes: Lifetime access

Course Description

The Hidden Markov Model or HMM is all about learning sequences.

A lot of the data that would be very useful for us to model is in sequences. Stock prices are sequences of prices. Language is a sequence of words. Credit scoring involves sequences of borrowing and repaying money, and we can use those sequences to predict whether or not you’re going to default. In short, sequences are everywhere, and being able to analyze them is an important skill in your data science toolbox.

The easiest way to appreciate the kind of information you get from a sequence is to consider what you are reading right now. If I had written the previous sentence backwards, it wouldn’t make much sense to you, even though it contained all the same words. So order is important.

While the current fad in deep learning is to use recurrent neural networks to model sequences, I want to first introduce you guys to a machine learning algorithm that has been around for several decades now - the Hidden Markov Model.

This course follows directly from my first course in Unsupervised Machine Learning for Cluster Analysis, where you learned how to measure the probability distribution of a random variable. In this course, you’ll learn to measure the probability distribution of a sequence of random variables.

You guys know how much I love deep learning, so there is a little twist in this course. We’ve already covered gradient descent and you know how central it is for solving deep learning problems. I claimed that gradient descent could be used to optimize any objective function. In this course I will show you how you can use gradient descent to solve for the optimal parameters of an HMM, as an alternative to the popular expectation-maximization algorithm.

We’re going to do it in Theano and Tensorflow, which are popular libraries for deep learning. This is also going to teach you how to work with sequences in Theano and Tensorflow, which will be very useful when we cover recurrent neural networks and LSTMs.

This course is also going to go through the many practical applications of Markov models and hidden Markov models. We’re going to look at a model of sickness and health, and calculate how to predict how long you’ll stay sick, if you get sick. We’re going to talk about how Markov models can be used to analyze how people interact with your website, and fix problem areas like high bounce rate, which could be affecting your SEO. We’ll build language models that can be used to identify a writer and even generate text - imagine a machine doing your writing for you. HMMs have been very successful in natural language processing or NLP.

We’ll look at what is possibly the most recent and prolific application of Markov models - Google’s PageRank algorithm. And finally we’ll discuss even more practical applications of Markov models, including generating images, smartphone autosuggestions, and using HMMs to answer one of the most fundamental questions in biology - how is DNA, the code of life, translated into physical or behavioral attributes of an organism?

All of the materials of this course can be downloaded and installed for FREE. We will do most of our work in Numpy and Matplotlib, along with a little bit of Theano. I am always available to answer your questions and help you along your data science journey.

This course focuses on "how to build and understand", not just "how to use". Anyone can learn to use an API in 15 minutes after reading some documentation. It's not about "remembering facts", it's about "seeing for yourself" via experimentation. It will teach you how to visualize what's happening in the model internally. If you want more than just a superficial look at machine learning models, this course is for you.

See you in class!

Suggested Prerequisites:

  • calculus
  • linear algebra
  • probability
  • Python coding: if/else, loops, lists, dicts, sets
  • Numpy coding: matrix and vector operations, loading a CSV file
  • Be comfortable with the multivariate Gaussian distribution
  • Cluster Analysis and Unsupervised Machine Learning in Python will provide you with sufficient background

Tips for success:

  • Use the video speed changer! Personally, I like to watch at 2x.
  • Take handwritten notes. This will drastically increase your ability to retain the information.
  • Write down the equations. If you don't, I guarantee it will just look like gibberish.
  • Ask lots of questions on the discussion board. The more the better!
  • Don't get discouraged if you can't solve every exercise right away. Sometimes it'll take hours, days, or maybe weeks!
  • Write code yourself, this is an applied course! Don't be a "couch potato".


Introduction and Outline

3 Lectures · 20min
  1. Introduction and Outline: Why would you want to use an HMM? (05:05) (FREE preview available)
  2. Where to get the Code and Data (09:21)
  3. How to Succeed in this Course (05:52)

Markov Models

4 Lectures · 43min
  1. The Markov Property (07:34)
  2. The Markov Model (12:30)
  3. Probability Smoothing and Log-Probabilities (07:50)
  4. The Math of Markov Chains (15:12)

Markov Models: Example Problems and Applications

6 Lectures · 36min
  1. Example Problem: Sick or Healthy (03:27)
  2. Example Problem: Expected number of continuously sick days (02:54)
  3. Example application: SEO and Bounce Rate Optimization (08:54)
  4. Example Application: Build a 2nd-order language model and generate phrases (13:07)
  5. Example Application: Google’s PageRank algorithm (05:05)
  6. Suggestion Box (03:10)

Hidden Markov Models for Discrete Observations

19 Lectures · 02hr 54min
  1. From Markov Models to Hidden Markov Models (06:03)
  2. HMM - Basic Examples (08:04)
  3. Parameters of an HMM (07:00)
  4. The 3 Problems of an HMM (05:43)
  5. The Forward-Backward Algorithm (part 1) (16:59)
  6. The Forward-Backward Algorithm (part 2) (07:09)
  7. The Forward-Backward Algorithm (part 3) (07:19)
  8. The Viterbi Algorithm (part 1) (06:15)
  9. The Viterbi Algorithm (part 2) (15:05)
  10. HMM Training (part 1) (04:41)
  11. HMM Training (part 2) (10:22)
  12. HMM Training (part 3) (13:34)
  13. HMM Training (part 4) (13:17)
  14. How to Choose the Number of Hidden States (07:02)
  15. Baum-Welch Updates for Multiple Observations (04:54)
  16. Discrete HMM in Code (20:34)
  17. The underflow problem and how to solve it (05:06)
  18. Discrete HMM Updates in Code with Scaling (11:54)
  19. Scaled Viterbi Algorithm in Log Space (03:39)

Discrete HMMs Using Deep Learning Libraries

6 Lectures · 54min
  1. Gradient Descent Tutorial (04:31)
  2. Theano Scan Tutorial (12:41)
  3. Discrete HMM in Theano (11:43)
  4. Improving our Gradient Descent-Based HMM (05:10)
  5. Tensorflow Scan Tutorial (12:43)
  6. Discrete HMM in Tensorflow (07:28)

HMMs for Continuous Observations

6 Lectures · 01hr 00min
  1. Gaussian Mixture Models with Hidden Markov Models (04:13)
  2. Generating Data from a Real-Valued HMM (06:36)
  3. Continuous-Observation HMM in Code (part 1) (18:38)
  4. Continuous-Observation HMM in Code (part 2) (05:13)
  5. Continuous HMM in Theano (16:33)
  6. Continuous HMM in Tensorflow (09:27)

HMMs for Classification

3 Lectures · 16min
  1. Unsupervised or Supervised? (02:59)
  2. Generative vs. Discriminative Classifiers (02:31)
  3. HMM Classification on Poetry Data (Robert Frost vs. Edgar Allan Poe) (10:37)

Bonus Example: Parts-of-Speech Tagging

2 Lectures · 11min
  1. Parts-of-Speech Tagging Concepts (05:01)
  2. POS Tagging with an HMM (05:59)

Helpful Review

3 Lectures · 18min
  1. Review of Gaussian Mixture Models (03:05)
  2. Theano Basics: Variables, Functions, Expressions, Optimization (07:47)
  3. TensorFlow Basics: Variables, Functions, Expressions, Optimization (07:27)

Markov Models (Legacy)

3 Lectures · 16min
  1. The Markov Property (04:40)
  2. Markov Models (07:03)
  3. The Math of Markov Chains (05:16)

HMM Legacy Lectures

7 Lectures · 27min
  1. The Forward-Backward Algorithm (04:28)
  2. Visual Intuition for the Forward Algorithm (03:33)
  3. The Viterbi Algorithm (02:58)
  4. Visual Intuition for the Viterbi Algorithm (03:17)
  5. The Baum-Welch Algorithm (02:39)
  6. Baum-Welch Explanation and Intuition (06:35)
  7. How can we choose the number of hidden states? (04:23)

Setting Up Your Environment (Appendix/FAQ by Student Request)

2 Lectures · 37min
  1. Windows-Focused Environment Setup 2018 (20:21)
  2. How to install Numpy, Scipy, Matplotlib, Pandas, IPython, Theano, and TensorFlow (17:33)

Extra Help With Python Coding for Beginners (Appendix/FAQ by Student Request)

5 Lectures · 52min
  1. How to Code Yourself (part 1) (15:55)
  2. How to Code Yourself (part 2) (09:24)
  3. Proof that using Jupyter Notebook is the same as not using it (12:29)
  4. Python 2 vs Python 3 (04:38)
  5. Is Theano Dead? (10:04)

Effective Learning Strategies for Machine Learning (Appendix/FAQ by Student Request)

4 Lectures · 59min
  1. How to Succeed in this Course (Long Version) (10:25)
  2. Is this for Beginners or Experts? Academic or Practical? Fast or slow-paced? (22:05)
  3. What order should I take your courses in? (part 1) (11:19)
  4. What order should I take your courses in? (part 2) (16:07)

Appendix / FAQ Finale

2 Lectures · 08min
  1. What is the Appendix? (02:48)
  2. Where to get discount coupons and FREE deep learning material (05:31)
This website is using cookies. That's Fine