How to Prepare for a Data Science Immersive Course
I am not an expert in mathematics. I am not an expert in science or statistics. Honestly, I feel uncomfortable saying I’m an expert in anything. However, I can say that I am proficient at a few things: learning is on that list.
Studying, teaching and performing music has taught me many things, but learning how to learn has been one of the most valuable byproducts. As a professional musician, you have to adapt and learn on the fly. If unsuccessful, that could mean the difference in being hired or not. Preparation is paramount. In this post I’ll share some methods that have worked well for me and I hope that some may find this beneficial.
Deciding to apply and attend a data science immersive course was a daunting step for me as I have no previous experience in computer science or mathematics. A quick google search left me feeling intimidated at how much knowledge I would need to be successful in the field. Can an immersive course teach you all there is to know about data science in a few months? After watching countless videos on youtube of young professionals detailing the trials and tribulations of immersive learning life I gathered a few things: immersive courses move fast and can be incredibly overwhelming. I am currently enrolled in General Assembly’s Data Science Immersive and can empathize with the YouTubers notions. I determined, for me, preparation was essential.
“How do you eat an elephant?”
Cliche and macabre as it is, I find this adage distills the essence of efficient learning. Start small. Start now. Be consistent. But where does one start?
Break it down
1. Assess your situation
2. Create a plan
3. Implement the plan
Assess your Situation
What do I need to know?
If you’ve decided on or enrolled in a course the syllabus is a great place to start. If not, this post by Taesun Yoo is a good place to start. What are the main areas of focus? Programming in python, data analysis using pandas and machine learning with scikit-learn are usually first on the list of topics. Make two list:
- Topics that you plan to learn throughout your course.
- Topics that you want to explore and familiarize yourself with before the course.
For me, both lists were long and my knowledge was limited.
Depending on the course, you may not need to know much beforehand. General Assembly does a great job at preparing you before the start of the course but additional study only makes the ride smoother.
Create a plan
Identify the basics. This is admittedly debatable. For me:
- Basic Python
- Data Analysis with Pandas
- Linear algebra
At this time, I whittled down my preparation list to a few topics: python, pandas, statistics, linear algebra, supervised machine learning, unsupervised machine learning, natural language processing and neural nets.
I decided to invest 80% of my preparation time in the “Basics”. This will be different for each person, obviously. If you have a PhD in statistics, you might not need an online statistics refresher. The rest of the topics on my list fall in the “Exploration” category where I spent approximately 20% of my preparation. While familiarity in the latter subjects are advantageous, you can’t get far without knowing how to define a function in python.
Udemy has great courses on python. Codecademy has a wonderful data science “career path” option that dives deep into python and SQL and has a high-level overview of machine learning and beyond. Kahn Academy is a great resource to brush up on statistics, calculus and linear algebra. Grant Sanderson’s YouTube channel, 3Blue1Brown offers a stunning and insightful series: “Essence of Linear Algebra”.
For “Exploration” topics, I attempted to immerse myself in articles and literature on Medium and TDS Team. Additionally, Lex Fridman’s podcast is a wonderful ride into the world of machine learning and AI.
There are countless other resources out there but these are the ones I utilized.
Implement the plan
Schedule your day. Decide how much time are you going to spend on preparation and segment accordingly. I decided on 4-5 hours per day (split into 2 sessions) for about 6 weeks.
Create reasonable deadlines for yourself. Most of the online courses give approximate timings for course completion which I used as a starting point. Typically, I found they significantly overestimated the amount of time needed to complete a course.
You may notice you need more or less time on a given subject after deciding on a specific timeframe. Adaptation is crucial. Personally, I needed more time understanding the concepts of linear algebra. I decided to spent 30 minutes every morning for two weeks attacking my problem area. Try to adjust your plan rather than abandon and start from scratch.
Table it. Take a break. Get after it.
I have a website bookmark folder with a long list of tabled topics. I find this necessary to declutter my mind. Some topics I came back to immediately and others are still waiting to be unearthed.
This is my simple, yet effective, approach to preparing for a data science immersive. In fact, I adapt this approach to learning anything.
I will continue to shape and adjust my learning style as I have throughout my education. Finally, I’ll recap the most important parts to me:
- Start Small
- Start Now
- Be consistent
Thank you for reading and I hope there was something in this post that helps you in your journey to becoming a rock-star data scientist!
Alex Fioto | Home
Schedule, contact, upcoming performances, and news for Raleigh and Durham-based trumpeter Alex Fioto