Data science is worth it

What do you think of data science courses?

Digit2  📅 27.05.2018 19:30:36
What do you think of data science courses?
Hey

Big data and data science are currently two big buzzwords, especially in the USA a lot seems to be emerging in this area.

While looking for the (hopefully suitable) course, I noticed that more and more data science courses are emerging in this country too.

What do you think of these courses? Is it worthwhile to specialize in this "trend" as early as the Bachelor's degree?

Best wishes,
Digit2



Edited 1 time. Last on 05/27/18 7:32 PM.
DS-Studi  📅 07.06.2018 12:09:01
Re: What do you think of data science courses?
Hi,

I am currently in a DS Master (after doing my BA in Info).

From Digit2 What do you think of these courses?
Opinions are now divided on this. Many say that DS is just an entertaining trend, for others it is Next big thing. The future will tell what is true of this.
I mainly opted for the DS Master because I wanted to get away from the classic IT career path and have always flirted with mathematics. The course as such is nothing special. Basically just a new combination of old math and information subjects. New tools / programming languages ​​are touched upon in every subject, but it is up to you to deal with them in depth so that you are "useful" for the economy. An affinity for programming is definitely helpful.

From Digit2 Is it worthwhile to specialize in this "trend" as early as the Bachelor's degree?
In my opinion: NO! The bachelor's degree should cover broad basics. You can specialize enough later and who knows which skills will be in demand in a few years. Maybe then the next IT pig will be driven through the village. If the area interests you, stick to the classic courses (math, physics, info.) And then keep looking.
Artificial idiots  📅 07.06.2018 20:57:04
Re: What do you think of data science courses?
From Digit2Is it worthwhile to specialize in this "trend" as early as the Bachelor's degree?
If you mean "doing data science as a specialization / specialization in a classic bachelor's degree", I would recommend you. The techniques from DS / ML are used more and more in a variety of industries. In 10 years' time, DS / ML is likely to become a kind of tool that all academics * must * master, just like MS Office is for today. Those who can do DS / ML are sufficient, but those who cannot do DS / ML at all allow themselves to be excluded from the job market. In any case, if you want to get a good salary, you will learn techniques from DS / ML sooner or later. It's smart to start with your bachelor's degree.
But if you mean "to do a bachelor's degree in data science", I would advise against it. The techniques from DS / ML are used more and more in a variety of industries. In 10 years' time, DS / ML is likely to become a kind of tool that all academics * must * master, just as MS Office is for today. Those who can do DS / ML are sufficient, but those who cannot do DS / ML at all allow themselves to be excluded from the job market. It doesn't entirely make sense to study a tool on bachelor's. Better to do something hard core like computer science, statistics or mathematics and switch to a DS master later.
Re: What do you think of data science courses?
From Digit2 Hey

Big data and data science are currently two big buzzwords, and a lot seems to be emerging in this area, especially in the USA.
A lot has already arisen, the hype was about 2 years ago. Back then, people were still looking for real specialists in the USA; today every Hans and Franz is a "data scientist".
I recommend studying computer science with a specialization in statistics / stochastics.
But in principle the basic modules Analysis I and II, linear algebra, statistics I and II and a little stochastics are sufficient.
You will learn the rest through books and learning by doing.
Be warned here, because of the GDPR you will have to learn a lot about data protection.
If you want to do data science, you should aim for the USA / China. China in particular is a paradise for such people!

BG
Re: What do you think of data science courses?
@DS-Studi Where do you study data science? I'm also currently looking for a computer science / data science degree (master's).
noDataScience  📅 16.04.2021 22:29:50
Re: What do you think of data science courses?
It is good that statisticians and mathematicians have noticed that computer science skills and programming are necessary in times of big data and the like. The same applies to computer scientists, who cannot operate serious machine learning without in-depth statistical knowledge.

However, the data science courses in general and especially those in DE do not reflect this. The term data science is not protected and it feels like anyone who has taken a module in quantitative methods can call themselves a data scientist.

Personally, I think that data science and the whole field of A.I., machine / deep learning is a lot of hot air. You just have to take a look at the data science or machine learning literature. One finds: Buzzwords that describe nothing else than linear and logistic regression models, cluster analysis, principal component analysis, random forests, maximum likelihood, ..., etc. That is pure statistics.

Data science is nothing more than applied statistics, with a little stochastics and some programming with Python or R or basic computer science terms. Sounds better than statisticians, computer scientists or mathematicians in 2021.

My humble tip: Keep your hands off it and do your bachelor's degree in computer science, mathematics or statistics and specialize in statistics for the master’s.
Re: What do you think of data science courses?
So, if you already know that you want to work in the field, nothing speaks against it in my opinion. But I'm also more on the side that you should learn the basics as broadly as possible and specialize in the master’s degree at the earliest for a very simple reason: You often decide to do something different. That is why studying computer science or statistics certainly offers more options.

Now on the subject of job prospects and the like: It is certainly true that data science will also play a role in the future, but you are increasingly noticing that the buzzword is losing its shine. The problem is also that the best times were around 3 years ago and there are many more applicants than jobs today. There are loads of white lateral entrants (many with a PhD) who apply for entry-level positions. Due to the popularity, the competitive situation is really enormous, we get at least 10 times as many applicants for one position as for e.g. developer positions and the clue is that significantly fewer data scientists are needed than you might think.

The next problem is the gap between academic data science and corporate data science. Frankly, statistical modeling is only a very small part of the job and many are very disappointed with the boring side of the job. Most of the time it revolves around cleaning up data, understanding any problems in the data. Modeling is also getting easier with tools like AutoML. And last but not least, the exaggerated expectations.

If that doesn't put you off, ok, but otherwise I wouldn't specialize so early.
Re: What do you think of data science courses?
From SWEngineerIt is certainly true that data science will continue to play a role in the future, but you are increasingly noticing that the buzzword is losing its shine. The problem is also that the best times were around 3 years ago and there are many more applicants than jobs today.

The next problem is the gap between academic data science and corporate data science. Frankly, statistical modeling is only a very small part of the job and many are very disappointed with the boring side of the job. Most of the time it revolves around cleaning up data, understanding any problems in the data.
You speak to me from the bottom of my heart
There are now really armies of kagglers who seem to have modeled every conceivable tensorflow-compliant pre-chewed data set, but when it comes to doing something from data sources actually available in practice (let alone collecting new data), it bites long in advance.
noDataScience  📅 17.04.2021 16:24:10