Evaluation of an Open Source Online Course from Stanford Engineer

Introduction

The course I decided to evaluate is an open source online course offered from Stanford Engineering, http://www.ml-class.org/course/video/preview_list ,  even though the course is coming from an engineering department, the course is called Machine Learning.  I used to be a Mathematics’ major before I switched to Psychology, so this will help me to truly evaluate the course since math as a lot to do with computer programing that can lead to AI- Artificial Intelligence.   These online videos are based on linear algebra.   The pre-requisite skills are the Algebra I & II, Geometry, and Calculus I, II, & III.  The ideas of Linear Algebra are a different way to solve problems with multiple variables that leads to computer coding.  When looking into the course you will hear words like matrices, vectors, and linear equations.  Now let us see how they planned and developed this course for distance learning.

Description & Planning of Course

The title of the course is Machine Learning; here there are a set of 3 sections called Introduction, Linear regression with one variable, and Linear Algebra review, which the last one is optional.  All of the sections consist of videos to watch all within a certain length of time.  The longest video runs for 14 minutes and the shortest video is 6 minutes.  Within the introduction there are 4 videos available that goes in linear order starting with the first video called Welcome.  Below is how the sections are set up.

I.                    Introduction

A.      Welcome – running 7 minutes

B.      What is Machine Learning- running 7 minutes

C.      Supervised Learning- running 12 minutes

D.      Unsupervised Learning- running 14 minutes

II.                  Linear Regression with one variable

A.      Model Representation- running 8 minutes.

B.      Cost function-running 8 minutes

C.      Cost Function- Intuition I-  running 9 minutes

D.      Cost –Function- Intuition II- running 9 minutes

E.       Gradient Descent- running in 11 minutes

F.       Gradient Descent Intuition- running  in 12 minutes

G.     Gradient Descent for Linear Regression- running in 10 minutes

H.      What’s Next- running in 6 minutes

III.                Linear Algebra Review

A.      Matrices and vectors- running in 9 minutes

B.      Addition and Scalar Multiplication –running in 7 minutes

C.      Matrix Vector Multiplication- running in 11 minutes

D.      Matrix & Matrix Multiplication- running in 11minutes

E.       Matrix Multiplication Properties running in 9 minutes

F.       Inverse and Transpose running in 11 minutes

 Each of the videos follow the same layout, you will see a person sitting in front of his computer screen, he turns toward the camera and begins to talk about the  topic having the video zoom towards the computer screen.  Then as you look on to the screen of the computer you will no longer see the person talking but you will be able to hear him.  The person talking (the voice) is always talking about the topic you see on the screen and when the screen starts to show a problem to work on, the voice then describes how to solve the problem using as many visual cues as possible.  Below is a picture of one screen with the voice describing the procedures on how to solve the problem.

                                                                         Figure 1

As you can see in Fig. 1,the voice goes through the problem he lets you know which number he is dealing with by highlighting it with a certain color, and h when he moves along the problem and to another number he changes the color of the highlighter, so when you want to pause the video you can see the connection of which numbers are used.  This procedure is great, because you can see how to solve the problem exactly as if you had a tutor in front of you as a F2F set up.  Another edition that is great is the practice function.  When you look back on the picture you can see two yellow hash lines on the video timeline, these markings let you know when a practice exercise is available.  When you get a chance to work on the problem you get a choice to skip or submit your answer.  There would be 4 answers to choose from, only one is correct. When you pick your answer and click submit you will be told at that point if it’s correct or incorrect.  If incorrect you can still work on the problem to get it correct, or if you don’t want to you can still have the choice to skip it.   There is no place on the video for you to write the problem out, you will still need a piece of paper and a writing utensil.  During the video clip playing you have the choice to press previous video or next video and to zoom in on the screen 1x, 1.2x, and 1.5x.

Conclusion

This open source online program was very well planned with visual cues to follow along and activities for the learner to participate (Simonson, 2009).  Allowing the screen to be legible with multiple options of increasing the screen which would make the letters and numbers bigger (Simonson, 2009).  The lengths of the videos are short to keep the student interested in the topic (Simonson, 2009).  Giving the student options to skip the problem or to see the videos multiple times gives the student options when taking the course (Simonson, 2009).  I enjoyed this course layout so much I gave the link to one of my students in high school who is thinking about electrical engineer for study, now he gets a chance to see the topics and get more information about the subject.  He was able to follow along and answered the problems from the Linear Algebra review.   

Thank you Andrew Ng for making this open source online course. 

References:

Machine Learning Retrieved October 6, 2011 from http://www.ml-class.org/course/video/preview_list

Simonson, M., Smaldino, S., Albright, M., & Zvacek, S. (2009). Teaching and learning at a distance: Foundations of distance education (4th ed.) Boston, MA: Pearson.