Reflective Essay about a Learning Challenge

This is an essay for the assignment of the course Learning How to Learn on Coursera.

Describe your current learning career and potential trajectory (if known).

I am an ex-financial analyst. 10 years ago I graduated from the Mathematical Methods in Economics Department of the Economics Faculty at a big university and for 7 years I pursued the career in investment banking, corporate finance, private equity and similar fields.

3 years ago I left my reasonably successful career in an office in a big financial center and moved to a sleepy touristic village on the sea coast of Spain where the only way that I could apply the skills that I'd been taught all my life was by finding a remote job. This life-changing event coincided with a kind of a burning out related to my previous job and I wasn't keen on returning to this industry. Therefore, I started to look for ways to enter in the remote/freelance Internet job market where I could apply my knowledge and skills, and my passions.

My passions are mathematics, analytical thinking, and languages, which were always my hobby. I'd been always learning languages at my free time for pleasure and it helped me very much when I moved to Spain because I had to reach a high level of understanding and using in 3 languages (French, Spanish and Italian) in rather short time. When I investigated job market I found that the remote and freelance jobs related to corporate finance are hard to come by (there are many reasons for this, but the main one is that the confidential nature of information used in this job requires corporations to employ exclusively office workers). But I didn't want to stop being an analyst, I liked that.

So I decided to switch career and become a data scientist that works with an open data, and I was very much keen on applying my language skills in this field as well. I decided to become a specialist in natural language processing related to learning and teaching languages.

Briefly describe the learning aim that is of importance to you.

The Data Science is a very exciting and innovative field. There is a lot of hype around this discipline, and it is so new that you are literally swamped with information. It is also quite an interdisciplinary field, requiring you to learn and combine knowledge from many fields, the large part of them coming from statistics and mathematics. The last challenge is that it requires a reasonable training in programming and understanding of computational systems. It is sometimes very frustrating that everybody thinks that you cannot enter into it without an IT background.

The challenges for me have been so far and continue to be the following:

• As I am learning online, from the MOOCs and books and practical assignments, I've had to organize my curriculum myself and follow it and analyze what is working and what is not and keep changing it on the go all by myself. I do not have a mentor or a teacher or an advising expert apart what I can find on the forums, chats, slack and other social communities on the Internet. With the amount of information there is at the beginning I was really lost and it seems to me now that I did make some mistakes that I could have escaped from. Anyway, by now I've been learning almost 6 months with some breaks and I feel that I've overcome my initial bewilderment.
• The second challenge is that after the university, even though I graduated with honors and excelled in the mathematical part of the curriculum, I've been applying my mathematical and statistical knowledge in the corporate finance field very lightly. It is not because of negligence, it is the requirement of the field that you have to be more skilled in the purely economic and accounting matters, like getting the information from financial reports, analyzing the stock exchange market, world and country economies, applying the methods of fundamental analysis to companies etc. While it all can be incredibly challenging and complicated economically, it is not really a challenging work from the mathematical and statistical point of view. The simple methods such as linear regression will suffice in most of the cases. Therefore, when I started to go through the necessary mathematical and statistical curriculum, I had to learn it all over again (though, of course, my university education turned out to be of an incredible value).
• My last challenge was my non-coding background. Being a financial analyst did not require an extensive technical knowledge of computers, databases, algorithms, programming languages etc. Proficiency in Excel and PowerPoint was really enough in most of the tasks. Some technically aspiring analysts learned how to program basic automation tasks on the Excel platform (using VBA language), but in the field it was considered even a bit way too geeky and generally was not necessary. When I started my Data Science curriculum in the January 2017, I was really a rookie in the computational and programming field.

For the purpose of this essay, I would like to concentrate on the last learning challenge as it becomes more and more persistent barrier to my further development. I do not have a coder's mind, and I still struggle with many things that would seem very easy to an IT professional. The biggest challenge I found is that you have to organize your time dedicated to learning in quite a different manner comparing to what I was used to before when I was learning mathematics and economics.

What is your biggest mental challenge in achieving your learning aim?

My biggest learning challenge turned out to be applying theoretical knowledge about programming and computing (as well data processing, data mining, data storing etc) to real work and assignments of the higher level of difficulty. Programming is a bit like sport and a bit like learning a foreign language in that you have to practice A LOT to become automatic in applying what you learn; you also have to practice and extensively analyze the work of others in order to understand what structures and expressions, what rules and vocabulary of programming languages you have to use in the different real-life situations.

If there is one field where you have to learn by doing, that is the one. And even though I haven't had such a rough time understanding complicated mathematical and statistical concepts, it's been really hard for me to learn to apply it in the real computational problems. And the most complicated things for me are the basics: reading files, working with databases and strings, Python data structures and their use, control flow constructs, regular expression and so on.

Furthermore, the usual learning path under supervising of the mentor in the office was completely unavailable to me due to the circumstances.

Outline existing research or learning techniques that are relevant to your challenges.

Deliberate practice. Deliberately focusing your studies on the more difficult material which you still have to master, to balance the illusions of competence which can result from repeated practice of the things you’ve already mastered. (For a complete discussion of this important concept by a key researcher in this area, see Ericsson, KA, and R Pool. Peak: Secrets from the New Science of Expertise: Eamon Dolan/Houghton Mifflin Harcourt 2016).

Chunks and chunking. A package of information made up of separate items which you have united in your memory through association or meaning. For example, a musical chord on the guitar can be a chunk—several chords can be combined together to become a larger chunk. An individual dance move might be one chunk—several movements combined together to form a larger chunk. Ultimately, an entire well-practiced dance can be conceived of as a chunk. In mathematics, one step in solving a problem can be a chunk. Several such chunks can be combined into a larger chunk which comprises the entire problem-solving strategy.

What we in LHTL refer to as a “neural chunk” is the neural pattern that corresponds to given chunk of information. See the excellent research of Alessandro Guida, Fernand Gobet and their colleagues, (for example, Guida, A, et al. "Functional cerebral reorganization: a signature of expertise? Reexamining Guida, Gobet, Tardieu, and Nicolas' (2012) two-stage framework." Frontiers in Human Neuroscience 7 (2013): 590; and Guida, A, et al. "Functional cerebral reorganization: a signature of expertise? Reexamining Guida, Gobet, Tardieu, and Nicolas' (2012) two-stage framework." Frontiers in Human Neuroscience 7 (2013): 590.) “Expert on expert” Anders Ericsson calls neural chunks “mental representations.”

Chunking is the act of grouping or organizing lists of information or concepts into compact packages of information that are easier for your mind to access.

Interleaving: Mixing up the types of problems or situations you are working on so that you need to switch between different techniques or strategies. (For a detailed explanation, refer to Spacing and Interleaving of Study and Practice, page 5 as well as Taylor, K., & Rohrer, D. (2010). The effects of interleaved practice. Applied Cognitive Psychology, 24(6), 837-848)).

Apply your knowledge of research findings or learning techniques to overcome your challenges.

The biggest part of my learning problem seems to be that I do not form steady and established chunks in programming knowledge before I move on to more complicated material. When I try to apply my flawed knowledge with insufficient chunks to the real life problems I bump into the basics that I haven't fully mastered and it starts to be really frustrating because the simple operations that I have to implement on the way to producing a bigger analysis or solving a bigger problem become the biggest obstacle in the whole task. I can see the solution to this problem in paying more attention to basics and studying them more deliberately and consistently, not breezing over them hoping that I will learn on the way. For example, I had a big problem with the regular expressions (the basics and their implementing in Python): after I spent 3-4 hours practicing writing regular expressions in the interactive programming simulator, I feel much more confident in using them.

Which leads us to the next the most important technique that I have to implement: deliberately practicing the material that seems to me challenging and difficult. Before I was more attracted to consuming mathematical theory and solving mathematical/statistical problems because it seemed to me less challenging than practicing coding. When I encountered programming assignments in my Data Science curriculum I tended to overuse the ready solutions that are presented on the Web (not each and every time of course, but I could do that in the most complicated situations). Now I see that I have to concentrate specifically on coding and using programming languages as the more easy understanding of mathematical and statistical part of material give me illusions of competence in this field.

The last technique that I think is really important for me is interleaving. As the data science problems tend to be incredibly interdisciplinary and heterogeneous, I have to learn to switch between different projects and different tasks from different fields, practicing the whole range of very diverse skills, where many of them have nothing to do with mathematics. It allows me to better understand the material because I do not have to skip parts of it because I did not study it yet as well as not tire so much of cramming one concept in my head where I can switch to something completely different but no less important.