What Study Do You Need To Become A Data Scientist? 4 Things I Did!

What Study Do You Need To Become A Data Scientist? 4 Things I Did!

 



This article explains recommended study methods for beginners who want to become data scientists.




Points of this article

  • Skills required to become a data scientist
  • Study method to acquire skills
  • Recommended study method I actually did

 

When you quit your career and take on the challenge of another career, you need to pursue studies that are directly linked to the career you want to become. The popularity has grown over the last few years, as has data scientists, the most popular profession for college graduates in the United States.


It is not widely known in Japan, so many people may not know what kind of study should be done to actually become a data scientist.


I will explain what you actually need to study and how much time I spent learning, so please refer to it if you are starting to study.


What is the Study Content Required to Become a Data Scientist?


Math

Data scientists need knowledge of mathematics because they work to analyze vast amounts of data and apply the results to corporate strategy. The actual analysis is carried out using machine learning, but knowledge of linear algebra, calculus, and probability statistics is indispensable for that machine learning.


If you originally graduated from a university in science, it may not be so difficult to review because you are in the first or second year of university, but if you are not studying in liberal arts or university, it may be a little difficult. ..


Programming skills

When data scientists analyze data, they need to process very large volumes of data, so relying solely on human hands and eyes is not enough. Therefore, as described in the section on mathematics, we will analyze using machine learning.


Learning a programming language is essential because machine learning needs to be modified to match the data you want to analyze. Specifically, the language Python is widely used for machine learning, so you should concentrate on learning.


Domain knowledge

Many people may not understand what they are saying even if they say domain knowledge. Somehow it feels like a URL? I will imagine it.


Domain knowledge is the knowledge and trends of the industry that works as a data scientist. How do you want to use it when analyzing large amounts of data? If you do not clearly understand this point, you will not be able to perform appropriate analysis.


By having the knowledge of the industry to which it is applied properly, you can demonstrate your abilities as a data scientist.


Study Time Required For Data Scientists

Now that you know what you need to study to become a data scientist, how long does it actually take to study? This depends greatly on how much original knowledge, experience, and skills you have.


Also, absorption capacity varies from person to person, so it cannot be said unconditionally. This is what is said at TECH:: CAMP, one of the major programming schools.


Q: I want to acquire the technical ability to create web services (tabelog and services like Amazon) by myself. How much study time do you need in total?


A: You need to have a total of 100 hours of study time. This is a good time to reach the stage of completing the basic curriculum and starting to create a prototype of the service.


This example assumes a web application, but data scientists should be prepared for at least this much.


The background varies from person to person, so let's make up for the lack of math, programming, and domain knowledge introduced earlier. We recommend that you start by clarifying the shortfall.


Of these, programming is the most time-consuming. If you are inexperienced, you will need to take the time to study, so if you are self-taught, it will take hundreds of hours. However, programming is also unsuitable, so the amount of study time required varies from person to person.


I think mathematics will take the next time. Even in science, many people are not good at probability statistics, and it takes more than 100 hours to study from the high school graduation level.


Domain knowledge varies depending on the company you work for, so the prior study should be enough to understand the entire industry. Dozens of hours are enough to look at multiple industries.


How to study data science that I actually did

Learn with free online materials

Especially regarding programming, online teaching materials have been enriched in the last few years, and free teaching materials are increasing as trial versions of paysites. I used it to get a sense of what kind of content I should learn.


Coursera 

Since it is a site for learning artificial intelligence programming, there are many Python courses used for machine learning among many programming languages. You don't need to build a computer environment and it is completed online, so you can get started easily.


Since it emphasizes practice rather than theory, it has the great feature that it can be advanced even before acquiring knowledge of mathematics. There is an introductory course on Python, so I learned how to use the library here.


EdX 

Compared to EdX, Courseerais not focused on artificial intelligence or machine learning, it is a site where you can learn a wide range of programming languages. You can take many courses by applying for the paid version, but even if it is free, the basic part can be suppressed.


This is also completed online, so there is no need to build an environment and it is easy to get started. What I found useful was that I was able to make effective use of the gap time while I was out because I was able to proceed with learning in collaboration with my smartphone. I was able to improve the efficiency of learning considerably.

Thanks for reading Subscribe Technology Illusion

Post a Comment

1 Comments