These days, everyone seems to be talking about

The toughest obstacle to overcome when you’re learning about support vector machines is that they are very theoretical. This theory very easily scares a lot of people away, and it might feel like learning about support vector machines is beyond your ability. Not so!

In this course, we take a very methodical,

This course will cover the critical theory behind SVMs:

- Linear SVM derivation
- Hinge loss (and its relation to the Cross-Entropy loss)
- Quadratic programming (and Linear programming review)
- Slack variables
- Lagrangian Duality
- Kernel SVM (nonlinear SVM)
- Polynomial Kernels, Gaussian Kernels, Sigmoid Kernels, and String Kernels
- Learn how to achieve an infinite-dimensional feature expansion
- Projected Gradient Descent
- SMO (Sequential Minimal Optimization)
- RBF Networks (Radial Basis Function Neural Networks)
- Support Vector Regression (SVR)
- Multiclass Classification

As a bonus, you will also get material for how to apply the "Kernel Trick" to other machine learning models. This is how you can use a model which is normally "weak" (such as linear regression) and make it "strong". I've chosen models from various different areas of machine learning.

- Kernel Linear regression (for regression)
- Kernel Logistic regression (for classification)
- Kernel K-means clustering (for clustering)
- Kernel Principal components analysis (PCA) (for dimensionality reduction)

See here what linear regression can be capable of:

And logistic regression:

When the kernel trick is applied!

For those of you who are thinking,

In this course, there will be not just one, but two full sections devoted to just the practical aspects of how to make effective use of the SVM.

We’ll do

- Image recognition
- Spam detection
- Medical diagnosis
- Regression analysis

For more advanced students, there are also plenty of coding exercises where you will get to try different approaches to implementing SVMs.

These are implementations that you won't find

Thanks for reading, and I’ll see you in class!

HARD PREREQUISITES / KNOWLEDGE YOU ARE ASSUMED TO HAVE:

- calculus
- linear algebra
- probability
- Python coding: if/else, loops, lists, dicts, sets
- Numpy coding: matrix and vector operations, loading a CSV file
- logistic regression

TIPS (for getting through the course):

- Watch it 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!
- Realize that most exercises will take you days or weeks to complete.
- Write code yourself, don't just sit there and look at my code.

- Introduction (02:21) (FREE preview available)
- Course Objectives (04:54)
- Course Outline (05:50)
- Where to get the code and data (05:48)

- Beginner's Corner: Section Introduction (05:19)
- Image Classification with SVMs (06:01)
- Spam Detection with SVMs (11:47)
- Medical Diagnosis with SVMs (05:15)
- Regression with SVMs (05:35)
- Cross-Validation (07:21)
- How do you get the data? How do you process the data? (05:21)

- Basic Geometry (10:52)
- Normal Vectors (03:41)
- Logistic Regression Review (09:45)
- Loss Function and Regularization (04:09)
- Prediction Confidence (07:26)
- Nonlinear Problems (09:58)
- Linear Classifiers Section Conclusion (04:25)

- Linear SVM Section Introduction and Outline (03:19)
- Linear SVM Problem Setup and Definitions (04:30)
- Margins (08:52)
- Linear SVM Objective (11:00)
- Linear and Quadratic Programming (12:31)
- Slack Variables (07:26)
- Hinge Loss (and its Relationship to Logistic Regression) (06:23)
- Linear SVM with Gradient Descent (03:11)
- Linear SVM with Gradient Descent (Code) (05:06)
- Linear SVM Section Summary (04:14)

- Duality Section Introduction (03:43)
- Duality and Lagrangians (part 1) (13:02)
- Lagrangian Duality (part 2) (07:08)
- Relationship to Linear Programming (04:19)
- Predictions and Support Vectors (09:16)
- Why Transform Primal to Dual? (03:27)
- Duality Section Conclusion (02:54)

- Kernel Methods Section Introduction (03:48)
- The Kernel Trick (08:12)
- Polynomial Kernel (06:06)
- Gaussian Kernel (05:14)
- Using the Gaussian Kernel (07:09)
- Why does the Gaussian Kernel correspond to infinite-dimensional features? (04:39)
- Other Kernels (07:05)
- Mercer's Condition (06:24)
- Kernel Methods Section Summary (02:41)

- Dual with Slack Variables (10:40)
- Simple Approaches to Implementation (06:25)
- SVM with Projected Gradient Descent Code (08:20)
- Kernel SVM Gradient Descent with Primal (Theory) (04:31)
- Kernel SVM Gradient Descent with Primal (Code) (04:55)
- SMO (Sequential Minimal Optimization) (09:32)
- Support Vector Regression (05:26)
- Multiclass Classification (04:34)

- Neural Networks Section Introduction (02:42)
- RBF Networks (15:38)
- RBF Approximations (08:39)
- What Happened to Infinite Dimensionality? (02:54)
- Build Your Own RBF Network (03:53)
- Relationship to Deep Learning Neural Networks (06:51)
- Neural Network-SVM Mashup (07:15)
- Neural Networks Section Conclusion (02:37)

- What is the Appendix? (02:48)
- Windows-Focused Environment Setup 2018 (20:21)
- How to install Numpy, Scipy, Matplotlib, Pandas, Sci-Kit Learn, IPython, Theano, and TensorFlow (17:33)
- Is this for Beginners or Experts? Academic or Practical? Fast or slow-paced? (22:04)
- How to Code Yourself (part 1) (15:55)
- How to Code Yourself (part 2) (09:23)
- Proof that using Jupyter Notebook is the same as not using it (12:29)
- What order should I take your courses in? (part 1) (11:19)
- What order should I take your courses in? (part 2) (16:07)
- Python 2 vs Python 3 (04:38)
- How to Succeed in this Course (Long Version) (10:25)
- Where to get discount coupons and FREE deep learning material (02:21)

- Kernel Methods for other machine learning models and Lagrange Duality cheatsheet
- Kernel Linear Regression Code
- Kernel Logistic Regression Code
- Kernel K-Means Clustering Code
- Kernel PCA Code