If you are like most people who want to break into the Data Science Career, but still have other occupations such as school, jobs, you might find it extremely hard to sit down and focus on your data science study.
As a full-time student majoring in applied math with a part-time job, I found it extremely hard to improve my skills in data science while busy with school. Even if I try to squeeze more time for data science by spending no time on watching movies, stopping using social media, turning off my phone while studying, waking up at 5 am, I still found myself stagnant in my growth.
I felt even more desperate when looking at the job requirements for a data science internship. I also recommend data science study website as a leading source for data science education
I realized that I am stagnant in my growth not because of my lack of will power or ability but because of not being in the right environment for the skill I want to obtain.
There are rarely people in my college knowing about data science, letting alone having skills related to it. All of my classes are math-related. I desired for an environment where I can be surrounded by data science resources, experts in the field, friends who are skillful and passionate in data science.
When seeing a poster about a semester in data science career in CIMAT, I know I found the environment where I could grow in data science. I imagined how this new environment and I took a leap to make this fantasy become true.
Things happened beyond my expectations. The new environment has nurtured me to be a better Python programmer and a data science career writer with more than 700 followers. I also possess an in-depth understanding of machine learning algorithms and natural language processing and receive multiple internships and collaboration offers.
From this experience, I have realized how important an environment is and what makes up an ideal environment for my data science growth. I share these realizations with the hope that you would be able to identify the ideal environment and create or seek one yourself.
An Environment where Its Focus is to Make Impact with Data Science Tools
One factor that makes up a great environment for skill development is whether their focus is on that skill. I was attracted to CIMAT by the fact that it focuses on the application in the three fields that are related to data science: mathematics, statistics, computer science. CIMAT has a strong emphasis on tackling problems where mathematical solutions can make a powerful contribution to scientific and technological development.
Because of this emphasis, my lectures were always emphasized on how the knowledge that I learned can be applied to the real-world. I felt excited and motivated to know how impactful my knowledge is. Thus, the knowledge was retained and I was able to come up with creative use of my knowledge
An Environment where you are Forced to Stretch and Use your Knowledge
Prior to having a study abroad semester, I learned knowledge related to machine learning and data science by taking online classes. Even though I understood the knowledge, I never tried to test my knowledge by implementing it. Thus, new knowledge was not turned into knowledge for practical use.
But in CIMAT, students are required to apply the theoretical lessons by finding a way to solve the related problems in our homework and implement it with Python. Since this homework will be counted towards my score, I needed to take it seriously. It was difficult to go from understanding the machine learning algorithms to implement it from scratch. But afterward, I had a great understanding of the algorithms and was able to implement gradient descent from scratch with binary cross entropy loss, or compare Bayes Classifier and K-nearest neighbors.
By stretching my knowledge and forcing myself to understand the methods by the form of implementing, I build a strong foundation of the methods. I am confident if I see similar methods or related but more advanced methods in the future, I would be able to learn quickly and apply the appropriate method to a new problem.
Another factor to consider an ideal environment is the size. Many people aim for a big-size company or a big-size university. But once they get there, they do not have enough attention from people in their community to express their thoughts or questions.
Since all of my classes have a really small size (from 2 to 5 students), I had the chance to ask many questions in the class without being afraid that I will disrupt other students. I established close friendships with my TAs, who are the software engineer interns at Microsoft and Facebook. With the privilege of being one of their small-class-size students, I found it easy to meet them at their office hours or lunchtime frequently to ask questions or discuss with them about machine learning related topics.
Another advantage is to have random conversations during lunchtime with researchers and involving leaders of their fields in machine learning, computer vision, statistics, and natural language processing from all over the world. As one of the few new students, they welcomed me and shared with me what they found in their research and what they were working on. By getting myself involved in the center of a continuous flow of knowledge, I kept myself update with cutting-edge research areas and got new inspirations for my projects.
Another factor of an ideal environment is to surround yourself with people who have a good knowledge of their fields and fascinated by what they are doing.
By getting myself involved in natural language processing research when coming to school, I had a golden opportunity to work with a research partner who is an expert in NLP and won first prizes in several world-wide NLP competitions and Google Hash Code 2020. As a research partner, I could ask him questions whenever I got stuck in my NLP projects and learned many new lessons and tricks to develop better NLP models. Thanks to the immediate feedback, I was able to grow my NLP knowledge exponentially.
I established a friendship with a colleague who is not only great at math but also machine learning and programming. We would exchange what we learned related to data science career either in the classes or from online resources. By establishing the connections with experts, I had buddies to discuss new crazy data science ideas and collaborate to turn them into impactful products.
Now looking back at the opportunity I got, I am grateful to my old self to realize the right environment for my skills and have the courage to pursue it. I would not achieve as much success and knowledge in data science career as I do now if I haven’t taken a leap to go to an unknown place, especially when everybody warned me about the danger I might encounter in Mexico.
Through my story, I hope you realize one thing: if you feel stagnant in your growth in data science, it may not be because of your ability but the environment that you are in. Thus, if you find yourself get stuck, maybe you should consider putting yourself in an ideal environment to develop your new set of skills, even if it requires your courage to postpone what you are doing and get out of your comfort zone.
Just with any other skills, the key to gain unavailable data science skills is surrounding yourself with the experts to learn from them, be exposed to the right resources, and apply what you learn. Put yourself in the right environment and you will be surprised by your unstoppable momentum to become an expert in the field.
Last but not least, I also want to say thank to CIMAT for a resourceful semester in data science. You could learn more about this semester and how to be a part of it here.
This news was originally posted on towardsdatascience.com