5 Vital Questions for a Mid-Career Change to Data Science

Data Scientist Dude
6 min readJul 22, 2020

--

Data science as a career is both interesting and rewarding. As more firms find ways to use data and create network effects there will be more opportunities for people who know who to manipulate data for profit. However, if you already have a job or career you may wonder if breaking into the field is right for you. You may have taken a few classes, love data and find the ability to predict preferences and forecast the future exciting, but is it really something you want to dedicate most of your working hours to master? Here are some important questions you may want to ask yourself and others.

  1. How Well Do I Know Math?

Not all of us were natural talents in math during our first time in school. All of us had that “oh-no-this-is-beyond-me” moment, whether it was in junior high Algebra or college-level Calculus III. Any subject related to computer science necessarily has math as one of its foundations. A person who knows arithmetic, algebra, trigonometry, calculus, and linear algebra will simply have an advantage, if only because they have developed a skill for problem solving. The good news is that today math does not have to be scary or boring. You are not limited to musty textbooks and dry examples. There are many free options, such as Khan Academy, or paid options, such as Udemy, for learning almost all levels of math online. Many of the most engaging teachers are the ones with the most views. MIT OpenCourseWare is of particularly high quality and free to use. Other fields that are necessary to know are logarithms, statistics, and general business domain knowledge. A foundation of even the basic concepts in these fields will make coding, constructing models, and cleaning data so much easier because you are not intimidated by numbers, variables, and constants. You do not have to be a theoretical mathematician and few people are but know that data science is rooted in computer science. Computer science is rooted in math.

2. Do I Want to Go into Debt?

There is an ongoing debate about whether “data camps” and online learning is better than learning data science in college setting. The correct way to think about it is to ask what you can afford. The quality of a comprehensive and skill-based education has skyrocketed in recent years. A person that wants to break into data science has several options today. A person can use a variety of private courses from firms such as General Assemb.ly, Data Camp or Coursera. Coursera is a not a bad place to start since it has a certain credibility being associated with Andrew Ng, one of the modern pioneers of machine learning and AI. All sources have excellent courses and classes that can teach you the fundamentals of programming and building models. The main advantages of going for a formal university education are that you will understand data theory better, have opportunities to network with fellow future data scientists, and have some sort of credibility built into your resume. If you want to teach full time or get a PhD this is clearly your path. Otherwise, keep in mind some of the advantages of paying for a university program can be gained through participating in hackathons and simply doing Kaggle projects on a regular basis. All told, it is better to start with what you can afford. You can best determine this by what you earn today, and maybe cut that total in half. Career changes are typically hard, and it will not be until you are an experienced data scientist that you may be earning six figures. Be realistic and build success slowly. After all, your greatest asset is really your computer and you will not have to go deep in debt to get a good one for data science.

3. How Do I Learn Best?

There are three basic ways that people learn. They learn from what they read, they learn from what they hear, and they learn from what they do. All of us favors one or two, and that is where you should spend most of your time. Data science is best learned by doing projects yet you are going to have to read or listen to something first in order to have some basic knowledge. If you extremely comfortable reading, then get one of many fine O’Reilly books on machine learning or some other data science subject and read it cover to cover, following along on your chosen IDE. If you are an excellent listener, then subscribe to a YouTube channel and follow along as they work through coding problems.

4. Do I Know an Expert in the Field That I Can Ask about It?

There are two ways you can answer this question if you feel stuck. The first is to use LinkedIn and send out notes to people with similar backgrounds with “data science” in their job description. If they decline to answer, be persistent by thanking them for their time and asking them if they could refer you to someone who can answer your questions. The second is to take an online course taught by a data scientist and ask them about working in data science full time during class or office hours. The best online classes I have had were with industry professionals who taught at night for extra money. They would often use stories from projects they were working on to provide an example. You can not only decide if data science is for you, but start to get a feeling for which industry you may want to work in.

5. What is My Authentic Self, and Can I Get Paid to be Me?

Some days our “authentic selves” may feel like the guy or girl who plays video games for hours while eating junk food. While this may be a little true sometimes it is rarely true all the time. Most of us have interests that stir something deep inside and motivate us to know more. During our childhood you may remember that wise Sesame Street monster Elmo ask, “Where can Elmo learn more?” These interests are probably closer to your authentic self. There is a real enthusiasm about the subject that feels natural. If you like sports a lot then perhaps working with data analytics and professional sports is something you would enjoy. If you simply like visuals, art, and computer design more than the prospect of using Tableau to communicate your findings may seem exciting. The important thing to keep in mind is that there must be a market match between your skills and problems that need to be solved. You have to create value in order to get paid, but the more the problem interests you the more likely you are to enjoy your work.

This fact is sort of a corollary to number five, but a bonus tip is that when it comes to choosing your career don’t blindly follow a passion. Many people will tell you to pursue your passion, and if you do that for a living and you will never work a day in your life. The truth is more complicated than this. Many people who have done just this, such as those working in humanitarian aid or education will tell you that honestly it really means you. Just. Never. Stop. Working. It is not as if you can shut it off at the end of the day (as you could do with your computer and cell phone if you are so inclined) and enjoy your free time. Some better advice is to be true to your authentic self, and double down on your talents and strengths no matter what the industry or firm. The auto parts store chain needs data science as much if not more so than your local bank. If you enjoy data science you will still solve problems, and likely develop a good technical reputation in your firm. Save passion for your personal life and give freely of your time to your family and charity.

The field of data science is an exciting one that offers many different opportunities if you are making a mid career change and find it intriguing explore all your options first you are never too old to learn new things. Simply put, you can do it because you wouldn’t be the first. Many other people have made a mid-career change to data science and are happy with their decision. Stay positive and plan on sticking with it many aspects are difficult to understand and they simply take time to comprehend learn and master. Nobody ever really stops learning in data science, no matter how many years of experience they have. Good luck.

--

--

Data Scientist Dude

Technology Strategist, Linguist and Autodidact - My mission is to help people understand and use data models.-