How to read course titles, for beginners

June 30, 2020

In this article, I hope to dispel some misconceptions about course titles, and to explain the convention I use.

This is a rare mistake made by beginners, but common enough to warrant an article.

The basic format used often is “General Subject: Specific Subject”.

The common mistake made by beginners is assuming that the course covers everything under the general subject, while not reading the specific subject, or perhaps not knowing what the specific subject is about (in that case, please ask using the “Contact” form above).

Let’s look at some examples.

 

Data Science: Deep Learning and Neural Networks in Python

In this case, ‘Deep Learning and Neural Networks’ are a specific instance of a topic under the general field of ‘Data Science’.

The title does not imply that the course covers everything about Data Science (such as hypothesis testing, data cleaning, visualization, dimension reduction, decision trees, etc.)

In fact, although “Deep Learning” is the “specific” subject here, this still does not imply the course covers everything about deep learning.

This particular course covers the fundamentals of deep learning, including deriving the backpropagation algorithm from scratch and implementing it in pure Python without the need for external libraries to do the heavy lifting.

Of course, that big long sentence you just read wouldn’t make a good title. 😉

As always, if you’re unsure, here are things you can do to get a better sense of the course:

  • Read the course description
  • Read the lecture titles and section titles (what I’ll refer to as the “curriculum”)
  • Watch the preview lectures, especially the introduction and outline where I describe what the course is about

Altogether, these should give you a good sense of what the course is about.

If, after doing all these steps, you still have additional questions, use the “Contact” form above.

I am happy to answer any questions and clear any doubts you may have.

 

Bayesian Machine Learning: A/B Testing

Again, in this case, the general topic is “Bayesian Machine Learning”, but the specific topic is “A/B Testing”.

This means that this course does not cover everything about Bayesian Machine Learning (e.g. variational inference, how to use PyMC3, MCMC, Gibbs sampling, ICA, MDS, etc.)

In fact, MCMC and Gibbs sampling are not even Bayesian topics – many people get this wrong. Solution: just ask!

So again, I hope it is clear that the course covers one specific area of Bayesian Machine Learning, which is the application to A/B Testing. Again, the pattern goes “General > Specific”.

 

Deep Learning Prerequisites: The Numpy Stack in Python

Another great example.

No, it does not cover every single prerequisite you need for deep learning – that is the “general” part.

The “specific” part specifies that this course is about the Numpy stack only.

One common mistake from beginners is that they assume it will cover every prerequisite they need in order to be prepared for deep learning.

What they often don’t realize is that what they need is like 2 years of education, which of course would not fit into a single course.

On average, I’ve seen that the beginners who tend to have the most problems start a bit below the “high school graduate” level.

In this case, you would require typical STEM courses at the high school level or community college level first (calculus, geometry, trigonometry, basic programming, physics would be helpful but not completely necessary, etc.)

Once you have done that, then look into the typical “core” curriculum taught in most STEM degrees in the areas of statistics, computer science, engineering, etc. These are:

  • Calculus 1, 2, 3
  • Linear Algebra
  • Probability
  • Statistics (nice but not 100% necessary)
  • Computer Science or Programming (with a bit of data structures)

Normally these are taught in year 1 and 2 of undergrad. I’ve discussed them here: https://lazyprogrammer.me/cs-degree/

 

Deep Learning: Advanced Computer Vision

Clearly, with 15 deep learning courses, not every course whose title starts with “Deep Learning” covers everything about Deep Learning. This course covers the parts of Deep Learning specific to Computer Vision.

 

Deep Learning: Advanced NLP and RNNs

Same idea again. General (Deep Learning) to Specific (NLP and RNNs). Clearly, if you only read the “Deep Learning” part, then there would be no difference between this and the last course! Which of course is absurd.

 

 

I think this is enough to understand the main idea, but of course, I could go on (I have ~30 courses at this time). Look at those if you want to see more examples.

Now you must be wondering: did I just invent this convention out of thin air? Is that why it’s so confusing?

No!

Here are some examples from my collection (links included so it’s clear I didn’t just make these up!)

 

Modern Robotics: Mechanics, Planning, and Control

Does this book cover everything about robotics, like how to use CAD? No. That’s the “general” part. The “specific” part is “Mechanics, Planning, and Control”.

If you go into the table of contents, you can see such chapters as: Rigid-Body Motions (mechanics), Forward Kinematics (more mechanics), Motion Planning (planning), etc.

 

Computational Biology: A Statistical Mechanics Perspective

Does this book cover everything about computational biology? No! There’s nothing in here about HMMs, phylogenetic trees, etc.

Computational biology is the general part. The specific part is “Statistical Mechanics Perspective”.

Chapter 1 is called “Equilibrium Statistical Mechanics”. Other topics include “free energy” and the “Poisson-Boltzmann equation”. You can look these up yourself to confirm that these are statistical mechanics concepts.

 

Biological Sequence Analysis: Probabilistic Models of Proteins and Nucleic Acids

Same thing here. Does this book cover everything about “biological sequence analysis”? No! that’s the general part. The specific part is “probabilistic models of…”

And in fact, a large chunk of the book is about Markov chains and HMMs (which are probabilistic sequence models).

 

 

I could go on, but I think you get the idea.

Now does this mean every course title uses this format? Of course not!

Examples:

Time Series Analysis, Forecasting, and Machine Learning

Financial Engineering and Artificial Intelligence in Python

I hope this article was educational and cleared any misconceptions you may have had about how to read course titles. As always, if there’s something you find confusing, use the “Contact” form above and you will get a quick response!

Go to comments


Deep Learning and Artificial Intelligence Newsletter

Get discount coupons, free machine learning material, and new course announcements