
The one thing everyone in the industry can agree on is that Data Science is hard to learn.
Like learning to golf, you could spend every day on the data science golf course that is coding, mathematics, and industry knowledge, and still not have mastered the game by the time you die.
Not only that, but some of the most complex concepts of data science are even enough to make the brave among us turn back.
Luckily, those who have come before us have left us a treasure trove of methods that we can use to take on the most complex topics to the point where we can become experts in them. The best method to use among them?
The Feynman Technique.
What is the Feynman Technique?
The Feynman Technique was developed by Nobel Prize-winning physicist Richard Feynman, a pioneer in the field of quantum computing and nanotechnology, who was known as the "Great Explainer" for the incredible lectures he delivered at Cornell and Caltech.
Despite being named one of the top 10 physicists of all time by the British journal _Physics World_, Feynman thought of himself as "an ordinary person who studied hard."
What made Richard Feynman the brilliant man he was, is the way he identified the things he didn’t know and then threw himself into understanding them. Over many years, Feynman offered insight into how he broke down the most complex topics of his field and distilled them into simple pieces of knowledge that anyone could understand. These insights have been collected and refined into the process we now know as "The Feynman Technique".
The Feynman Technique is a four-step process for understanding any topic.
Instead of relying on basic recall achieved from rote memorization, the Feynman Technique favors developing true comprehension of a topic through active learning.
The Feynman Technique can be broken down into four steps:
- Choose a concept to learn.
- Teach it to yourself or someone else.
- Identify your knowledge gaps and return to the source material.
- Simplify your explanation.
How the Feynman Technique works.
When it comes to learning the complex topics of data science, it’s highly likely that you think you understand something until someone asks you to explain it to them or when someone asks you to apply it to an analysis.
The Feynman Technique solves this issue because it doesn’t give you the chance to believe that you’re a master of a subject when you’re just an amateur. Instead, it forces you to evaluate your knowledge at each step in the process, identify any knowledge gaps, engage directly with the material to fill those gaps, and then clarify and refine your understanding of the knowledge.
Why use the Feynman Technique?
The Feynman Technique is the perfect methodology to use to learn the complex concepts of data science because it forces you to confront what you don’t know and then guides you into a realm where you know it forwards and backward.
Learning the complex topics of data science is a daunting task. This is further compounded if you believe that your learning experience is complete after reading an article about the topic you are trying to learn. Remember: reading is not understanding.
Instead, by using the Feynman Technique, you ensure that your understanding of a concept is solidified.
The key part of the Feynman Technique that makes it so effective is the simplifying step. Famously, while preparing for his upcoming undergraduate lectures at Caltech, Feynman realized that he couldn’t explain the topic at an undergraduate level. At that moment, Feynman realized that he didn’t fully understand the topic if he couldn’t explain it to a classroom full of freshmen. This led him to completely rework the topic until he had simplified his explanation and created analogies that were appropriate for undergraduate students, thus leading him to a deeper comprehension of the topic.
Using the Feynman Technique to learn the most complicated concepts in data science will bring you to a point where the topics are no longer scary or complicated because you have simplified your understanding of them to the point where they become clear and easy to understand.
How to use the Feynman Technique to learn the most complicated concepts in data science.
Step 1: Choose a concept to learn.
Selecting a topic to study compels you to face what you don’t know and face it head-on. Once you’ve selected a topic, take out a blank sheet of paper or open a new Word document. Title it with the name of the topic.
Then, write down everything you want/need to know about the topic. This could include understanding how it works, how to apply it to analyses, and when it should be applied to analyses.
At this point, try to write down everything you already know about the topic. Don’t go looking on the internet to get you started. Using your own brainpower is what creates an effective learning experience. This is also the point where you should try to use existing knowledge to understand the new concept.
Make a note of any point where you must use a large complicated word to explain something, then ask yourself if you understand what the word means. If you do, great! If not, break down what the word means and then try to figure out a way to describe the process using simple language.
Step 2: Teach it to yourself or someone else.
This is the critical step that will tell you how much you understand the concept. Teaching it to yourself or, ideally, someone else, makes it very hard to trick yourself into believing that you are an expert on the topic.
Teaching the concept to yourself or someone else will force you to refine and simplify your explanation to the point where even a 10-year-old could understand it.
Teaching also initiates a feedback loop where questions and critiques will help you sharpen your understanding of the topic. Welcome questions, and write them down if you don’t know the answer yourself. This is a great way to help you identify further gaps in your knowledge.
The benefit of teaching the topic to yourself or someone else is that it builds your confidence. When this happens, you will be more inclined to take on tougher topics and you will have the confidence to apply what you have learned going forward.
Step 3: Identify your knowledge gaps and return to the source material.
This step is part of an iterative cycle with Step 2.
More often than not you will identify knowledge gaps after you first attempt to teach a concept to yourself or someone else. When this happens, you must return to the source material to answer questions, refine your understanding, and solidify what you’ve learned. Remember to keep writing down your understanding onto that sheet of paper started during the first step and to use simple language when adding new key points.
Once you’ve gained a new or further understanding of the concept, you can go back and teach the concept again. This cycle should be completed as many times as is necessary until you know the topic forwards and backward.
Step 4: Simplify your explanation.
The final step involves simplifying your explanation or understanding of the concept to the point where you can explain it without using jargon.
Jargon is often used when people don’t understand what they’re talking about, either because it’s an easy way to explain a topic without putting much effort in, or because they just want to appear like they know what they’re talking about.
Because of this, this step is the most important in the process. Forcing yourself to explain the concept using plain language doesn’t allow you to hide behind big fancy words. Instead, it forces you to truly break down the concept into bite-sized pieces that can be understood by anyone.
Additionally, now is a great time to make simple analogies for the concept, which are both easier to recall and explain in the future.
This is also the time when you should simplify any code you’ve written while working on a concept. Often we will write extremely complex code the first time around when understanding a new concept. Now is the time to write the simplest, cleanest code possible that is easy to understand but still gets the job done.
The takeaway.
Learning data science is hard. Period.
The Feynman Technique is the perfect method for learning the complex topics of data science because it favors developing true comprehension of a topic through active learning. This method of learning ensures that you can then apply the concept in your real-life data science work.
Here are the four steps of the Feynman Technique that you need to remember:
- Choose a concept to learn.
- Teach it to yourself or someone else.
- Identify your knowledge gaps and revisit the source material.
- Simplify your explanation.