Building training learning and loss
The effective control of the learning and building process should be the real motivation to continue reading the following text.
I would like to introduce myself just a bit.
I recently received my Ph.D. in technical cybernetics. My web of science H-index is around 5, which is not that bad actually. I was interested in control theory, machine learning related to image recognition, time delays, running, I love my current job, and many more.
Effective control of loosing
Beautiful control theory can be applied to a various range of science disciplines, engineering, project management, learning process, and real life. It is called control theory for a reason. It can be applied to many problems and shape your senses towards better decisions.
This is what is constantly happening to me “ building, learning, training and loosing”. Do you know this process? It doesn’t matter what kind of learning, neither building or training. The losing is coupled with all the progress forward. It is naturally frustrating. Yes, indeed it is.
The question is formulated as follows: how can I effectively control these cycles? Let’s dive into this so common problem from the natural point of view.
Motivation examples
I passed lots of math courses in my life. I am constantly learning new things in computer science, engineering, and losing my already acquired knowledge. I hate this. I am starting with linear algebra again and once over again until the rest of my life. The good part of this losing problem is faster learning of already seen topics. There are completely blind spots. There are islands of wisdom. There’s the frustration. What if I was born with a faulty brain?
The wintertime is just behind the front door and my running season is close to the end. I have reached great times on 10 km and 5 km this year. I am talking about 43 minutes per 10k. I will lose most of the progress during the winter for sure and start once again in spring. Luckily, I started my next season with a better set of tools, and goals.
I am playing guitar for almost 20 years. Is this going straight forward?
I did most of the college courses for completely nothing. I found myself comfortable just in a few of the many. At least, There was a goal.
Computer vision and machine learning were my main scope when we started a business for marketing and analysis of customer behavior from security cameras. It fails. There was just a lack of executive experience in general. The process of learning was pretty intensive. I am grateful for this failure more and more.
No matter if you are doing business, building an Instagram account, or learning new skills and knowledge. You can follow some common patterns to influence the process of learning and building.
Set the goal
If Arnold Schwarzenegger saying ”you have to have a goal because without that vision and goals you are drifting around and you never gonna end up anywhere”. What do you think? Is this true or false?
You cannot go forward if you are not aware of what you want. You can found an optimal strategy to achieve anything. You cannot optimize your personal “loss function” if you do not set the goal.
You cannot apply the best possible strategy without the gaol. There is nothing like, It will turn out somehow in the building process.
Vision
The awareness of the current standing and the goal allows you to pick up the right steps forwards. If your goal is known, by measuring how far you get, it helps you to prevent drifting around or nowhere.
Tools
Actions, the set picked up just for the all necessary step forward. You have to have a vision and goal to pick ut the right ones.
Can you imagine controlling the car just with pedals? What if your goal is just slightly right from your current direction. You are definitely missing something to reach your target. Maybe steering wheel.
Feedback
Evaluate your actions. You need to have a goal. You cannot evaluate how effective your tools are without the goal. Failure is just a mirror, just feedback to not repeat the wrong path.
If you are not progressing forward. You do not have the right tools or you do not evaluate the outcome correctly. It is simple like this.
If your Instagram account is not generating followers. Maybe the content is not interesting enough. Maybe the needed tools are different. The “interesting” content is produced by a million users and the right advantage (new tool) is to pay and advertisement content.
There are limitations. You should be aware of your limitations. Control theory talks about controllability and observability. When you can’t move the rock you may need dynamite. Sometimes, the uncontrollable fact cannot be changed, but sometimes new tools can change the game.
Lack of observability can make your life worse. How are you going to evaluate the outcome? Just the goal is not enough if you are not aware of the current state.
Robustness
Your optimality criteria can be of various types. It depends on what you want. Boring, gaol again.
The problem of robustness can be part of your optimality criteria. Control theory knows this. There is one more thing.
If you want to be robust, it decreases your ability to move in a certain vertical.
This principle is relatively simple. If you focus to be the best programmer or pole vaulter in the world. Your goal should be relatively narrow. Strong focus and time investment to be the best one in some areas can be your goal. You might have a different one. I want to be a good runner, programmer with a great level of guitar skill. I need a balanced resource between these activities and obviously there will be better ones than you. There will be better runners, programmers, and guitar players with a strong focus on just one of these activities.
You know what? There is a chance, to be the best one in this group of three activities.
Just set your goals. Sprint or decathlon. How universal you wanna be?
It is up to you. Do you want to be the best ballet dancer or your goal is wider than this vertical? Do you want to make the best cars? Let your company focus on cars.
Learn useful things you enjoy
It makes a lot of sense to creates a scenario, where the acquired theory is transferred into practice. The university is a place to discover a lot of new areas and options. The simple extending subconscious amount of knowledge has positive effects. This approach can bring your steps in front of many closed doors from different perspectives. The value of this knowledge will be dramatically reduced by time, but once you turn knowledge into a working prototype. You will barely forget.
Life after college is slightly different. It makes much more sense to focus on the knowledge areas applied to your work. You know what? It depends on your goals.
Last but not least in this chapter. Do what you like. True clichés.
Time-varying environment
Conditions can change. You have to react no matter what. If something that was working fine like feedback, actions, robustness strategy doesn't work anymore. You may already be playing different games under different rules.
Timing
This is so important topic. If you have all the mentioned in good shape. You can by applying actions move towards your goal (certain conditions). Just consider the initial states. Are all possible initial states equal to each other? Certainly not. It is like starting a travel agency in COVID 19 age. Consider this timing to perform the initial moves. Do you have some advantage in the pocket even the conditions are worse?
Inequality of sampling
Nyquist Shannon sampling theorem is a very important thing. You can observe various violations of this law. The right sampling to evaluate processes and progress is essential. The violations come from ignorance of the measured phenomenon or let's say system dynamics. This have serious consequences for evaluation.
Time constants
Be aware of time constants. This is as important as the others. Effective learning or building is tightly bound with the transient phenomenon. Just take into consideration the combined containers of energy and your effort. You need to know how frequently to apply your actions. You have to evaluate how long it makes sense to apply which action. Do you remember the weird graph in this article?
You can apply action or invest some effort into acquiring new knowledge or develop some kind of physical strength and end up with the red course.
You can go running once per month and end up with the orange line of the graph.
You can apply the actions or running a training session wisely and follow the green line.
This is a kind of important fact about building. You can reach tiredness while learning or training. You need to be aware of time constants for how long to take some rest, for how long to apply the action, and when to apply again. What is the period between action when all my progress is gone? What is the period while my brain forgot all the acquired knowledge? What is the time to recover and start once again?
Conclusion
I am focusing on effective learning, building, or growing while minimizing loss. I tried to bring control theory terminology into this. Our success can be influenced by many factors. What do you think? Are these things behind every path to success?
Goal setting and awareness - robustness - actions and the importance of the right ones - feedback - timing and time constants.
Can you minimize the loss better now?
