Early into my self-improvement journey, I gained the mindset of anything is possible, if I set my sights on it. There is a catch though, given enough time and effort. Which is very important detail because that is necessary for managing expectations. Life is all about managing expectations. If the expectation was to understand Linear Algebra or Calculus in a short period of time without having a firm grasp of the prerequisites. Then I was setting myself up for heaps of disappointment. However, if the expectation was that it would take a year or more of dedicated study to accomplish that task. Then I could start to see myself actually being able to accomplish that objective. This would be the foundational mindset for success. It would take a lot of time, work, and effort. Though I knew that I could do it.

Machine Learning

There was a realization one day that I might be a bit behind the curve. This happened after watching a presentation on what neat things can be done machine learning. At this point in time, I was still paying attention to what folks in the industry had to say about new developments in technology. I gathered a feeling that the future is moving towards a more data-centric and machine-learning enabled solutions. Wanting to remain in a position where I could still make contributions to that new future made me feel like I needed to learn a new skill set. However, there was this deep sinking feeling because I knew my math skills where awful. How could I expect to make machines dance with numbers if I lack a strong foundation in mathematics?

I was able to skate by college and university on a shoddy foundation of math created by neglect in my early learning days. A full re-build of my foundational skills felt like a prerequisite to have any semblance of success in these data science fields. At least, that is what I thought at the time. The only way forward that I could see was to re-learn all the basics. Then work my way back up to a level where I could do something interesting. In an attempt to instill the confidence I can attempt to do anything machine learning related.

At the base of the mountain there was a really long winding road up. This is where my new journey began. The measure for success was to be able to mostly understand, complete, and pass the exams provided by the MIT Linear Algebra course. Which would then instill the confidence that I could start taking all the courses at fast.ai. However, in order to reach that point I had to completely redo my foundational mathematical knowledge.

Human Learning

One of the very first tasks that needed to be done was to leave my education ego at the door. Essentially fully embracing that my foundational math skills where that of a middle-school student. Thankfully, there is an amazing resource that was built to help children learn maths online https://www.khanacademy.org.

Here is the learning track that I followed to get to a place where I felt comfortable taking the Linear Algebra course offered by MIT:

• Algebra I
• Geometry
• Algebra II
• Trigonometry
• Pre-Calculus
• Statistics & Probability
• Differential Calculus
• Linear Algebra

The resources provided by Khan Academy are put together really well and easy to comprehend. Performing the exercises before being shown how to solve them really helped build my mathematical intuition. The absolute mass of content needed to be consumed was the most difficult part about this leg of the journey. At the time, I had 4 hours in-between working, sleeping, and exercising, to work on whatever I wanted. Therefore, I dedicated the first two hours after waking up six days a week for the next year and some change to overcoming this task.

Part of this goal was to not finish as quickly as possible, the ultimate end result was to have a really strong fundamental understanding of math. Even though completing the required courses helped, understanding the content being taught was the true treasure sought. Math is a perishable skill, which meant that most of the topics being learnt would be forgotten as time went on. Therefore, what I wanted to retain was the confidence and necessary intuitions to hash out the problems again in the future.

That being said, it took an immense amount of willpower to accomplish this task. What kept me going on the tough days was the thought of not being intimidated by math problems. Another trick that helps is if you ever find yourself struggling to care mid-way through a large task, then you can do the following:

1. Step away
2. Consume a small amount of pure sugar (8 grams works for me)
3. Come back to the task 3 minutes later.

Our brains run on glucose, meaning the sugar essentially replenish the "Give a damn" tanks in your brain. You should find that you are able to hold your focus on the task at hand much easier than before. It may be placebo, but it works for any difficult task and not just math problems.

The difficulty and duration of this goal also prevented me from dedicating the entirety of my free time to reaching this goal. I wanted to remain hungry to complete the task. Meaning I wanted to still see the value in continuing. This was a monumental task that I thought needed to be completed at the time.

There were times that it felt like I was doing things inefficiently because of how I partitioned my effort. Like I was moving slowly to accomplish things because I was only doing half of the work. There were another 2 hours that were not being taken advantage of. What made things more difficult was the same thoughts started occurring with the tasks on the other side of the day as well.

I was moving slowly at accomplishing two things now. All because my attention was divided between both of them. However, I knew that I am not a math focus juggernaut and would stop caring about completing it if my entire time was dedicated towards this task. So looking back, I think this division that is what kept me going. I remained hungry to complete the task.

I finally reached the milestone where I could no longer put off starting the MIT Linear Algebra course. The goal for this course was to understand the concepts and not finish as quickly as possible. I wanted to follow the syllabus of the course and pretend like I needed to pass. Which would then give me the mindset to work on AI stuffs. To accomplish this I: watched the video lectures, then read the chapters outlined, worked on the assigned homework problems, and took the tests provided. While I am grateful these resources are provided for free the difficulty factor of this course was next level. Which made this a pain in the rear. Even grading the homework proved to be difficult because a teaching assistant was not readily available to help us out.

Thankfully, I found somebody else to work alongside me when completing the course. We spent the next 3 months (I think) spending our free time working the homework problems, completing exams, and studying together. We finally reached a point where we were able to take the final exam. I distinctly remember going into that exam and treating it like it was the real thing. The feeling of "Yeah, I understand what is being asked here" probably made up for the past year and a half of dedicated effort.

Dabbling in A.I.

Once we felt like we felt like we had our fill of difficult mathematics the next phase of education entered the picture, machine learning. https://www.fast.ai/ is the free resource that we used as the entry point into this world of data science. This portion of my maths journey was much easier to work on. Solving the problems in these courses actually yielded things that were interesting. Rather than ending up with just a notebook filled with notes, I actually had a program that did interesting things.

Even better was that I felt confident going into the courses and continued to feel that way as the courses progressed. I felt as though I understood the concepts at their core. In the end, I was able to completely go through all the material provided by the nice folks at fast.ai. Which included building a machine learning model from scratch.

After having consumed everything that was offered from this resource and hitting a major milestone, I distinctly remember the feeling of "Now what?" The goal to be able to work on machine learning problems became actualized. I was at a crossroads in my journey I was faced with a decision on what to do next. Having fulfilled the purpose of being able to understand the inner workings of a complex and foreign topic, I needed to know what was in store for me. Everything that I work on has be directed towards a goal. Now that I reached my destination, I did not know what to do with myself.

There were some options available. I could continue to learn more about various machine learning and data science topics. Various data science related challenges and programs also existed. All of which I could participate in. However, by this time I had convinced myself that I really did not care to get into the data science field. There were not any problems that were fascinating enough to continue to keep my skills sharp. It might have been that I was not looking hard enough, but there was not anything worth working towards at this point.

Late 2019, I decided that it was time for me to move onto the next interesting challenge. Looking back at things now, I realised there was not a need to complete all the work upfront to start to dabble in machine learning. However, I am grateful that I did do all that work, because I now have more confidence when approaching mathematical problems.

In the end, having three notebooks filled up with the notes of my learning was not all that I was left with. I also gained the assurance of the fact that I am capable of whatever I put my mind to. Even though the skills that I learned where perishable, the intuitions that I earned can easily be re-fortified. Who knows, maybe one day I can find an interesting problem to work on that enables me to take advantage of these higher level maths skills.

Until that day comes, I have my eyes open for any new opportunity that allows me to work on a fulfilling problem. Something that will not only benefit me, but the others that I share this world with.