Are You Covering or Discovering?
In this column I want to talk about three perspectives on learning.
Teachers are given the desired goal for the students to reach in the form cover material X during the period Y so that they are prepared for next course Z. Specifically, you are to teach this course and cover these chapters in the book. Your goal is to cover all of the material in that course so that the student does not get hurt in the follow-up course.
As a teacher since 1993, once a professor always a professor, I can tell you that this goal is often easier to achieve the first time one teaches because you are nervous and are so focused on achieving it. The second time you teach a course it seems that we want to get into the details more and emphasize the key ideas even more.
As a learner, you are on the receiving end of the challenge that the teacher is going through. You are moving so fast and trying to absorb at all to the best of your ability. In the end, what was covered and what you retain are seldom equal.
Lately, I have been re-learning machine learning. As I have written before, my doctoral work was in AI 94–99. I wanted to test my idea of mastery of a subject within the short timeframe that I wrote about. You might argue that the power of that article is that is to be applied to a new area you want to master and you would be correct. I am cheating a little.
As a re-learner of machine learning, I am finding myself falling in love with the ideas once again. It has been tough to tie the white belt and have a beginner mindset. BUT I am allowing myself the freedom to REDISCOVER the key ideas. I am not worried about timeframe but I am worried about filling in gaps.
Recently I even signed up for a Stanford course by Andrew Ng on the topic. Excellent course that COVERS many of the ideas from the field. I then take the time to put pen on paper and REDISCOVER the nuances that I thought I saw in his presentation.
What I have concluded is that instructors, courses, and books COVER subjects but that you must sit down and REDISCOVER the material for yourself in order to make it your own. The COVERING will provide you with 80% of the tools BUT the REDISCOVERY will fill in the rest of the 20% to make you complete.
I also have been working with several sources and seeing perspectives from different people. Some want people to learn the material, Andrew Ng, and are excited about the potential of what others can do with the material. Others want to impress their colleagues with their superior mathematical manipulations. Machine learning requires both BUT I lean towards Andrew’s work.
Impressing others with mathematical manipulations without pulling them along is useless except maybe for someone to get tenure. I am able to follow them BUT that is because of my math abilities. Writing at such a high level scares most people even if you COVER the material, the reader may never DISCOVER it due to it being inaccessible.
So for the non-math folks, teaching others requires the ability to translate complex ideas into something understandable. As a learner, we should take the time to digest that information and plan on spending time discovering the nuances that make that idea powerful.
My four cents…
