🎉 Data Science Transition - Week 1

Last updated May 15, 2022 Edit Source

• Data science
• Career

I’ve told many people in my life that I’m interested in data science, and made it pretty clear that I intended to get into it after I graduated from college. After many attempts, I still struggled to find a foothold in the field. However, that all is about to change.

This week I am sharing my first steps in my journey to becoming a data scientist. While exciting, this is but a small step of many. Data science is a huge field, encompassing everything from building dashboards and generating business reports to training large-scale machine learning models to intelligently solve problems at scale.

I don’t know what the end of this blog series will look like, but I know that it will not be the end of the road.

The main purpose of this blog is so that you - the reader - can hold me accountable in this journey. As author Tim Ferriss expressed in his book The Four Hour Body - and as have others in many other places I’m sure - sharing your goals with others and allowing them to track you is a great way to improve your odds of success in trying to reach a goal. Also, who knows, maybe someone else will find my journey useful.

I think it’s probably important to address why I’m deciding to do this. My first attempt to become a data scientist was primarily driven by the prestige and the pay that came with the title. The downside of this was that it was hard to convince myself to work on something that I saw as having little value, and when it was more difficult than I expected to break into the field I became discouraged.

According to Edward Deci and Richard Ryan’s work on self-determination theory, the predominant source of drive is not extrinsic motivation such as pay or prestige, but rather internal motivation. They found three key factors associated with intrinsic motivation:

• Competence - you feel improvement after putting the work in.
• Authenticity - what you’re doing reflects your core values.
• Relatedness - what you do involves cooperation and feeling a part of a community.

Having circled back to this career path after starting to make headway as a software engineer makes me believe that there I see aspects of competence, authenticity, and relatedness in this field. Indeed, this field is huge and is one I would like to get better at, reflects my core value of lifelong learning, and has a vibrant community online. As I explore this field more I’m sure I’ll be able to update those, but for now, I feel like I have sufficient intrinsic motivation to pursue it.

Pursuing intrinsic motivation is also valuable in and of itself. I think this is something that the modern educational system could be a lot better at. We only allow people to begin to explore their interests in earnest in college, by which time they have already missed so much time in developing the skill to do so. I believe learning to follow your curiosity and internal motivation is one of the most important skills in the modern era.

Jay McClelland - one of the fathers of deep learning - talks about how intrinsic motivation is crucial for skill development. Motivational engagement (read: nurturing intrinsic motivation) results in immersion (deep work), which then creates the opportunity to obtain the expertise.

$ \textrm{Motivational engagement} \rightarrow \textrm{Deep work} \rightarrow \textrm{Expertise}$

You can check out this Lex Fridman podcast to hear the original quote.

That being said, let’s jump right in and I’ll share how I’m going to conduct this transition.

My plan for this career transition was adapted from the following awesome video by Sundas Khalid. In it, she suggests having a targeted approach by starting with researching the job you would like before beginning your path.

This led me to create the following plan for myself:

1. Research the job market to target my desired job, and create a learning plan. 1 - 2 weeks.
2. Refresh concepts from statistics that are relevant to data science. Time: 1 - 2 months.
3. Join competitions and engage on Kaggle, an online data science community. Time: 1 - 3 months.
4. Create projects on Github to showcase my skills. Time: 1 - 3 months.
5. Prepare for interviews. Time: 1 - 2 months.

I’m mostly done with step 1, and have already started to review concepts from statistics that I think will be most relevant for data science and machine learning. More information on that will be coming soon in my next blog post.

If you would like to keep up with my journey you can enter your email below to subscribe. I’m planning to send out weekly updates on Sundays with my progress from the previous week. If not, that’s cool too. I believe my odds of success are higher just by putting this out on the web and telling people about it, and who knows - maybe somebody else could use this information as well!

I'm a freelance software developer located in Denver, Colorado. If you're interested in working together or would just like to say hi you can reach me at me@ this domain.