How do students learn best? What is the best way to learn?

As a current master’s degree student, I can say that the learning process in my bachelors was hard. It was hard not only because of complex subjects, but also the wrong method of teaching. I actually liked the way most of my professors used to teach. But the best was the project-based learning.

In this post, I will tell how do we, students, learn in the best way. I will also talk about the project-based learning and, why it is so effective. Hopefully, this post will be helpful to students and also to professors.

Learning new information and their dependent concepts:

Photo by Max van den Oetelaar on Unsplash

Let’s assume that the student will learn the subject “Image Matching” from computer science. Student must know several subjects and concept before he/she learns it.

Previously necessary information for the will-be-learned information. From most complex to most basic:

2-dimensional convolution and correlation(signal processing), representation of images digitally (signal processing), matrix operations (linear algebra), arithmetic operations, definition of number. (There may be other subject to be learned previously. These are the ones that I guess.)

First, let’s assume that the student is in second of year computer science bachelor’s degree. This lets us to assume that the student’s most complex previous knowledge is matrix operations. So, the student must learn 2 subjects in signal processing which are convolution and correlation while he/she also needs to learn the representation of images digitally. The convolution and correlation operations can be done in time domain and frequency domain. If we want the student to learn extensively, then we teach both of these subjects. But for frequency domain, student must learn calculus. So, I would skip the convolution and correlation operations in frequency domain so that we can reach to our goal (teach image matching) faster.

If we do not teach the convolution and correlation in time domain to student as a previous knowledge, student’s brain will relate the image matching subject to another concept which may cause disturbance in learning of the subject accurately. If this happens, feedback process will take longer time.

Feedback in learning:

Example: Student learns the subject 100%. Student is asked to explain the subject. He/she tells the subject by cause and effect relationship.

Case 1: One cause he tells for an effect is missing. So, the learned effect is wrong.

Case 2: One cause he tells for an effect is slightly or dramatically wrong. So, the learned effect is wrong.

Case 3: One cause he tells for an effect is the effect of another previous cause. The concept relation he/she made in his brain for previous cause is wrong. So, the cause is misunderstood, which then causes the last effect to be wrong.

When we learn a subject, we create dozens of cause and effect relationship in our brain. Even though the probability of a failure is so low, the effect of probability is pretty high. Because one small failure in the chain of cause and effect relationship may cause huge mistakes at the end.

How should we do the feedback in learning ?

After the investigation, wrong outcomes should be corrected and taught to student.

Figure 1- Relation to Previous Concept and Cause and Effect Relation

Why the project-based learning is best?

My own experience for project-based learning:

My image processing professor gave me a project to complete. The project is to implement an image processing method told in a scientific paper. I chose a paper which investigates focus measure operators. I had no idea about focus measure operators. These three words were completely alien to me. While we were having our lectures in image processing, I was also researching and learning about focus measure operators. They were basically edge detection methods. What I was learning in the lecture was also edge detection. I learned that the operations in time domain are done by filtering kernels (Example: 3x3 matrices). So, every mathematically weird looking method with all those weird mathematical symbols started becoming easier to understand. Because, I related the edge detection to a basic concept which is matrix. I was not seeing those formulas as weird mathematical symbols anymore. I was seeing them as moving, multiplying, subtracting, summing and dividing natural numbers.

I learned and implemented the methods in paper. I understood them all. Plus to that, I started seeing the deficits in those methods, because of concept relation I made. I created a new method to detect edges and measure the focus. The lecture’s professor and some other professors liked the new method I developed a lot.

Recently, I developed a machine learning algorithm that uses same filtering kernel concept. Surprisingly, nobody has done it before. When I tell it to people, they find it too complex. But to me, it is nothing more than moving and operating arithmetically numbers.

Thanks to my image processing professor Mehmet Türkan.

Tekin Evrim Özmermer

AI Developer In “>”, Interested in Artificial Intelligence, Human Intelligence, Economical inequality, and all other interesting stuff.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store