abeytheo data science journey

Let The Journey Begin!

I am elated to start this new year with my brand-new blog! It’s been a long time since I wanted to have a personal blog and finally, it came into realization. In a moment, I will tell my story how I got introduced to data science and my career in data field so far. I would also share my favorite resources to learn and keep up to date on this rapidly evolving field. So, let’s roll!

The Buzz Words

Even though research on AI and machine learning was not new at all, data science and machine learning have just been quite buzzing words in recent years. Computing power advancement was said to be the main reason for these subjects to resurface. In addition, if we look at Google trend below global interest has been increasing steadily since around 2014.

alt text

Career Shift

My first interest in machine learning arose around 2015. At that time I found out that statistics, linear algebra, calculus, machine learning algorithms, and programming are the rudimentary knowledge required. Those are a lot of requirements and actually impossible to master within a short period. At that moment I was doing a Technical Consultant job at a software vendor and my major weakness lies on the math and machine learning algorithm parts. So, I realized that I needed to find my way to learn the basics enough before committing a career shift. Then, I decided to allocate some hours after office to take a popular Machine Learning online course delivered by the one and only, Andrew Ng.

Before proceeding any further, I would like to give a brief review. This course is friendly to people with software engineering background as there are some math formulas but not too heavy. The course was delivered from an intuitive perspective with examples and it suits software engineers well IMO. Coupled with a few programming assignments in Octave (MATLAB alike), learners get the chance to see the actual code and study the algorithm implementation. All in all, this course is a helpful introductory course on machine learning.

At last, I was fortunate to grab a career as Data Scientist at Tokopedia. Nowadays, I highly doubted for a company to hire a Data Scientist simply by understanding the concepts without real-world experience or at least doing any projects.

My Experience

I developed few machine learning models and had experience deploying the models to production while working at Tokopedia. I definitely learned a lot there and maybe I will create a few posts related to my past experience.

Machine learning requires a lot of data to be collected in order to leverage sophisticated algorithms such as Deep Learning and also simply to have more features to experiment with. Based on this need, I got the opportunity to delve into crafting data infrastructure as a Data Engineer before finally decided to pursue Data Science field further by taking a Master degree at the University of Twente with StuNed scholarship grant.

Resources

I once heard this phrase said by my ex-CEO, William Tanuwidjaja

We are the luckiest generation ever. In the internet era, everyone, even the underdog has the chance to challange the status quo, stand against all odds, survive and eventually win.

And I found this quote to be ultra relevant in data science/machine learning field, in fact also generally in technology field. Nowadays there are abundant available resources on the internet one could benefit even without an academic degree. Finally, here are my favorite online resources:

  1. Khan Academy
    This website provides a wide variety range of machine learning fundamentals, including descriptive and inferential statistics, linear algebra, calculus, and also python
  2. Coursera
    Apart from machine learning course by Andrew Ng, there are plentiful resources revolving around Data Science here. Notable course is a deep learning course from deeplearning.ai, a company founded by Andrew Ng.
  3. DataSchool
    I use pandas a lot for data analysis. Kevin Markham delivered fun contents there, including some data preprocessing techniques and some machine learning concepts.

For more advanced stuff, I also like openly public courses taught by Stanford University:

You might be surprised, there are a lot of video tutorials available free in YouTube delivered by good content creators. Some of my favorites are:

Finally, to keep up-to-date with current machine learning development, I rely on Twitter by following scientists, organizations, and influencer in data science/machine learning. Here are some of them:

Some people might say one must understand all those maths, statistics, and machine learning algorithms before doing an applied machine learning. Quoting from Rachel Thomas, one of the founders of fast.ai, it is actually more feasible to do it with top-down approach instead. Pick one project, start coding, and learn the heavy stuff along the way.

Going Forward

Phew, there goes my first blog post. Please give me some feedback as I am thriving to improve :) I have been working on several projects for my university courses and there are also some interesting projects I have in mind. So, expect more exciting contents ahead. Thanks for your time and happy new year 2019!

Written on January 2nd , 2019 by Abraham Theodorus
Feel free to share!

You may also enjoy:

comments powered by Disqus