What a Month of Data Science Bootcamp has Done for Me
Introduction
The transition from an industry to Data Science is a unique and an individualistic process. People from different backgrounds can teach themselves data science through textbooks, online resources, and/or through peers. However, often this is not enough to become an established data scientist in industry, as best practices and expectations may not be fully informed to self-taught individuals. While it may not matter to various employers where one is taught on how to be an effective data scientist, it’s usually more credible to express how a guided and structured form of learning has helped somebody enter the world of Data Science. The individual passion for learning should never be discredited for its capabilities — each method of teaching cannot be successful without discipline and motivation to make a change. An apprentice of the field must always be open to discovery inside and outside of guided practice, as the soul of the field requires a following with strong critical analysis.
However, with all the above mentioned, we are going to particularly focus on one side of this spectrum by exploring the world of Data Science Bootcamps — the controversial yet effective educational model dominating the tech world as we know it today. In this article, we are going to feature the successful company founded in 2011, as an established, global presence in the market of Data Science education. We will discuss about how a single month of enrollment into the school’s demanding and full-time immersive program has made a change on an individual. To do this, we will receive a full dissection of a personal anecdote from none other than yours truly, the author….me!
The Pursuit of Education Through General Assembly
The first question we all should ask is, “Why Bootcamp?” Bootcamps are a great alternative toward obtaining a form of education for a subject in a cost-effective manner and in a relatively short amount of time. The tradeoff for pursuing a bootcamp instead of 2–4yr commitment from a credible university is the lack of a degree (legitimate bootcamps offer certification in place of a degree). To combat such tradeoff (among others), bootcamps, such as General Assembly, are often focused on getting their cohorts “industry ready” by graduation, focusing primarily on the tools and methods used heavily by many industry giants while simultaneously stressing the importance of due dates. One example of such skills taught is the ability to properly use Amazon Web Services (AWS), which is considered to be the leader in cloud computing across different industries for years. Furthermore, such bootcamps will introduce a bulk amount of information and work onto their students in an effort to have them drive hard into their studies constantly and completely envelop themselves in the field they are pursuing. Such demanding curricula prioritized on deadlines, allows bootcamps to successfully graduate dedicated cohorts ready for industry. Through effective time management and sacrifice, cohorts can learn enough to successfully step into their new professions.
What a Month of General Assembly has Taught Me.
Week 1
Week 1 of General Assembly did not shy away from the “nitty-gritty”. To keep up with the expanse of technology and its industry integration PLUS to keep up with the overwhelming amount of data generated for analysis, General Assembly spent its first week having their students understand the basic concepts found to operate Python (an open-source successful object oriented programming language meant to simplify “coding” for the non-technical audience). Python is a wonderful tool to not be underestimated in its difficulty of use. It’s a relatively easy language to start on, yet exceedingly difficult language to master. For the purpose of analyzing data, Python is a useful tool to help automate processes (as handling data by hand would be extremely cumbersome). Within week 1, students were able to dive into Python by navigating through the importance of functions, practicing control flow, and understanding data types (the ultimate, being a rather important foundation to recognize in any software tool!)
During this week, I personally had began to realize the sacrifice I needed to make to successfully complete this course. There was not much time to linger around in relaxation. The pace of the curriculum was quick, the information being retained was difficult, and the change was real.
Week 2
Upon successfully completing week 1, cohorts spent the following week understanding the world of open-source packages affiliated with Python. Packages such as “pandas” and “matplotlib” were introduced as staples in the language, widely used by all “pythonistas” (a silly word to describe coders who use Python). An importance on data visualization and data cleaning became apparent as well as an importance on recognizing data bias.
During this week, the change in routine began to sink in and begin its wear and tear on me. Projects, labs, and quizzes were due while in-class workload was increasing immensely. It was also here where I began to realize that the majority of my time will be spent cleaning data. Data was being provided in a relatively “clean” manner to us as cohorts, but these were the baby steps needed into understanding the daily challenges a data scientist will face.
Week 3
In the third week, we dove into some mathematical understanding of regression Machine Learning models. Machine Learning was taught to be one of the most important aspects of data science as the art of making computer models predict data accurately would be extremely valuable skill to carry into the real world.
I personally began to enjoy what I was doing much more as I explored what value modeling brought to the field. It was rather interesting to understand the importance of cross-validation — an effective way to better understand training models by analyzing subset permutations of data as a whole. I also found it interesting to see how features are engineered from other features (it was so intuitive to think about in practice and elegantly used simple logic to explore hidden data relationships).
Week 4
By this week, we finally moved into different types of modeling for supervised learning techniques. We were introduced to the vast word of classification modeling, by being exposed to KNN modeling, hyperparamters, logistic regression, and more.
To Sum It Up…
After a month in the course, I am happily able to say that I have improved tremendously since when I have first started. I have tried to teach myself data science on my own through some textbooks and other resources, but was unable to succeed on my own. In the 4 weeks of the course, I have learned so much and consistently apply what I am using throughout the entirety of my work. Classification modeling has been my favorite form of modeling thus far, and I hope to do more work with such models. I am extremely satisfied with the progress made and look forward to making more strides as all of the above is aforementioned is only the surface of what is to be learned and covered in data science. I sincerely believe that the rate at which the bootcamp moves will help me cover much of what is needed to keep up with industry. In the future, we will dive into web scraping, ensemble modeling, and SQL databases. I hope that many others do not discredit the value such bootcamps can bring to those willing to make the leap of faith.
