You might hear that Data Scientist is the sexiest job in 21th century. Moreover, Glassdoor’s recent research reveals that Data Science is the best job field when earning potentials, job satisfaction and number of job openings are examined. However, custom Data Science major still does not exist in bachelor degree. Furthermore, there is no concrete path to become a Data Scientist. So, how to become a Data Scientist?
In the most general sense, a data scientist is someone who is better at statistics than any software engineer and better at software engineering than any statistician.
Data Scientists have to be competent on mainly Math, Statistics, Programming and Database. Even though, Math is the weakest link for CS graduates among these fields, they are the strongest candidates in software developers. Because, almost 50% of software developers never received a degree in Computer Science. For instance, significant number of developers are employed from Electronic Engineering graduates. However, Electronic Engineering education is mostly fed from Physics whereas Computer Science education is mostly based on Math. Herein, having a CS background is a plus because math and statistic ability is a must to become a Data Scientist. On the other hand, I wouldn’t bet on non-developers to be strong about programming and database.
Data Scientists should already have strong programming skills. Today, scripting languages is in great demand for ML studies. That’s why, becoming familiar Python or R is reason for preference. However, you should not put up a wall around you with these programming languages. Ageless technology does not exist. You should be open to new technologies. In my bachelor years, classmates discuss about becomning Java developer or .Net developer better or profitable. I figure out that this discussion was really ridiculous later. Today, this discussion might substitute for being Python developer or R developer better in ML events. I am sure this discussion would keep on forever for different languages. That’s why, these technologies should mean rather than an end even if you are fan of a programming language.
You may also study some popular machine learning frameworks as a beginnig. Even though, these frameworks restrict you, they provides to produce faster solutions. Tech companies have run against to announce their machine learning framework in recent year. Herein, Google’s TensorFlow is ahead of the pack. That’s why, I have recently created an online training course about TensorFlow. I strongly recommend to be enrolled the course for beginners.
Academic background is also a plus for this field. You cannot find a CS related job that asking MSc or PhD degree in job openings except Data Scientists. You might think to have an academic career. In my opinion, you should at least have a MSc degree. Supportively, I always reap the benefit of having MSc degree.
Developers who does not have a CS background might be more convenient for Data Engineering. This title is in charge of DB administration, data storage and system implementation whereas Data Scientists responsible for building models and visualization. BTW, if you find a Data Engineer and make him Data Scientist, then you have got a superstar.
As mentioned, Data Scientists have to be able to build mathematical models, develop code. What’s more, this discipline requires to ask right questions and communicate with business people to get the right answers. In other words, they don’t have to be expert of some business domain knowledge. Being curious would get you rid off in most cases.
Solving linear solutions is expected from a standard software developer. In other words, they should moslty implement data manupulatin language scripts in backend. Whenas, Machine Learning Engineers should calculate partial derivative of a function, work on different optimization algorithms. In other words, being Data Scientist requires non-linear implementations.
Visualizing skills are another importatant requirement for Data Scientists. You may hear the phrase a picture is worth thousands of words. You might think visualization as PR. Because, people can only be convinced by when they see convincing materials. Additionaly, statistics is also important for marketing. There are three kinds of lies: lies, damned lies, and statistics. Indeed, facts might be abused or manipulated by ones who know statistics well. A well statistician can show underwhelming results as heartwarming.
To sum up, if software developers who have CS background would invest on math, then they could be cut out for Data Scientist. Non-developers should not lose hopes, they should invest both math and programming. So, you can build the future as a Data Scientist.
Hopefully, the post would contribute to develop careers for new Data Scientists.