By Kurtis Pykes, Machine Learning Engineer Intern
March 2020, I received a call informing me that I would be furloughed until further notice — informally meaning I’d be paid to learn. I knew the probability of me being made redundant after the furlough period ended was high since there were no projects I was actively working on.
Even though I hadn’t been doing much work with data at work, the thought of not being able to do any meaningful work with data bothered me. Nonetheless, I felt like my options regarding what I could possibly do next were limited since I did not get much practical experience at work. Don’t misunderstand me, I had been doing work as an intern, but I hadn’t done anything to significantly (or even marginally) improve the business (at least in my eyes) in my time. I was in a very low place, lacking self-belief, doubting my skills… For me, the furlough couldn’t come sooner.
The first transformative decision I made was to commit to becoming a future-proof indespensible Data Scientist.
When you make a commitment to do something, a force from within drives you. I wake up every day thinking I must be better today than I was yesterday and that is what drives me. However, for this post, I am going to share the 3 things I did during my furlough period to ensure that I move closer to my goal.
I was comfortable when it came to explaining theoretical concepts in Machine Learning, but I wasn’t satisfied.
Whenever I looked on Kaggle to see what solutions people were using, I’d always see some form of Boosting, Bagging or Deep Learning. Boosting and Bagging, I had a good understanding of, but Deep Learning was a no go zone for me. It was when I realized this, I decided to enroll in the Deep Learning Specialization on Coursera.
Learn Deep Learning from deeplearning.ai. If you want to break into Artificial intelligence (AI), this Specialization…
In this course, I learned many fundamental Deep Learning architectures and techniques to improve a Deep Learning model.
I was already pretty decent at programming, but whenever I listened to podcasts about how people have built their careers in the field, one thing always stood out to me.
Reinventing the wheel is good to gain a deep understanding!
I’d never coded a Machine Learning algorithm from scratch and this caused me to question whether I actually knew what was going on.
Consequently, I set a challenge to myself to code many of the most popular machine learning algorithms from scratch — for those that have been following my post for a long-time, you’d know that is the Algorithms From Scratch Series.
Algorithms From Scratch – Towards Data Science
Read writing about Algorithms From Scratch in Towards Data Science. A Medium publication sharing concepts, ideas, and…
Additionally, I thought it was essential I improved my skills with key Data Science frameworks such as NumPy and Pandas, hence I also created PyTrix Series.
Pytrix Series – Towards Data Science
Read writing about Pytrix Series in Towards Data Science. A Medium publication sharing concepts, ideas, and codes.
I decided to increase the frequency of my post from once a week to 3 times per week. This change forced me to do 2 major things which I believe has been pivotal to my growth:
- Constantly Learn
- Simplify and communicate what I’ve learnt
Constantly learning is imperative as a Data Scientist. We all know how fast technology is moving, so to remain sharp we must sharpen our axe. However, when you learn a new topic with the intention to regurgitate that information to someone else, although I have done no research about this, I find I absorb the information differently — I think deeper about what I am learning and try to picture it in my mind which all contributes to making learning a seamless process.
The best Data Scientist isn’t the smartest one.
Soft-skills aren’t taught in most MOOCs, you have to go out and learn them yourself.
The requirements to become a Data Scientist such as knowing how to program, statistics, linear algebra, calculus, and other key data concepts often consume the aspiring Data Scientist so much that it is easy to forget the most important part of being a Data Scientist… Being able to understand what the business wants to achieve and then use the data to add value.
In other words, a good Data Scientist knows lots of technical concepts but what separates them from the great Data Scientist is the ability to take a technical concept then simplify and communicate it in a manner that is inclusive of all members of the team, regardless of their technical level.
“If you can’t explain it simply, You don’t know it well enough”
In my personal opinion, every Data Scientist is a personal brand. The Wikipedia definition of Entrepreneurship is the creation or extraction of value — effectively, this is the essence of being a Data Scientist.
People usually only begin networking when they believe it’s time for them to land a Data Science role which I believe is complete nonsense.
Build your Network before you need it.
Here are 5 ways that building a Data Science Network has helped me improve in the past 8 months:
There are people that have gone ahead of you in life and I personally believe the universe allows us to cross paths with these people, so they can guide you. Let’s face it, at points of your Data Science career, you are going to need to ask for help!
A strong network is a great testing ground for ideas. I have conducted many polls on my LinkedIn which provides me with instant feedback. Additionally, you can get personal referrals. As things stand, I’ve never had to apply for a job because I’ve always known the power of word of mouth and I have used it to my advantage on numerous occassions — regardless of the field.
Learn From Others
You can not possibly know everything in Data Science (or about life in general) and having a diverse network of people will expose you to new things. Experience doesn’t matter, if you have a well-built network, you will learn something new.
If you ask the people I grew up with what Artificial Intelligence is they may respond with something from Black Mirror. Having no friends in the field can be quite lonely because there are definitely times when you will feel tired, unmotivated, and your non-Data Science friends probably would not be able to understand you. Networking with other Data Scientist will allow you to realize that you aren’t the only one in the world facing a certain challenge(s) and it is certainly what lifted me back up when I felt down.
People Know What You Have To Offer
The glue holding any relationship together is attached to both sides of what each person brings to the table and that is it. If people know what you do, it’s much easier to introduce you to someone else — this is how I landed a freelance gig in August.
A key thing to note is that I already had lots of exposure to the field which is what has allowed me to progress as I have; I’d say the most important thing I have done to change the trajectory of my career was to commit. Commitment is a long-term decision and bettering yourself daily is in only your hands. Taking responsibility for where your career is, is the beginning of developing yourself. Although I am nowhere near where I would like my Data Science career to be, I am closer than I was yesterday and much closer than I was 8 months ago.
Let’s continue the conversation on LinkedIn…
Kurtis Pykes – AI Writer – Towards Data Science | LinkedIn
View Kurtis Pykes’ profile on LinkedIn, the world’s largest professional community.
Bio: Kurtis Pykes is a Machine Learning Engineer Intern at Codehouse. He is passionate about harnessing the power of machine learning and data science to help people become more productive and effective. Follow his Medium blog: https://link.medium.com/2tFtAhN7d7.
Original. Reposted with permission.