Is data scientist a promising profession?

AI and big data are getting hot . Countless training classes began by speaking Python、 Classes like machine learning make a lot of money , The business majors in the University have begun to set up modeling and prediction courses . More and more resumes received in enterprises claim to have played neural network 、 Deep learning , experienced N Multiple modeling and forecasting projects . Future workplace , It looks like there will be data scientists all over the world .

From the perspective of the demand side, this trend does exist . The data accumulated slowly , The concept of artificial intelligence is also spreading rapidly , People are beginning to have this kind of consciousness , They all want to use the accumulated data to make analysis and prediction , This kind of business has gradually become a lot , Naturally, there is more demand for employees , Now generally speaking Data scientists are few and expensive It's still normal , somewhat AI It is not difficult to find a job with skills , The boss worries about you running away every day .

---

however , Data scientist as a profession , In the long run, it may not be very promising .

Why? ？

Because there is another field that has also begun to be popular with artificial intelligence , Namely Automatic machine learning software .

Like this , Yiming modeling, produced by Runqian, a well-known domestic data software company YModel, Genuine domestic products , It's completely free , Go to Runqian's official website to download . Let's feel it ：

Just get the data ready and throw it in , A model can be built in a few minutes for prediction , There is no need for people who understand data mining to intervene in the process , That is, we need to know some indicators of model evaluation when we finally look at the effect . All the trouble is data preparation , This is something automation software can't do , But it also has nothing to do with what data scientists are good at .

This is just one of the lighter ones , There are more software that can automatically do machine learning , Include google And other major manufacturers are joining this camp .

---

If you don't study the principles of these modeling algorithms carefully , I just learned some concepts and operations in training classes and crash courses , Then you may not be able to do this .

In the process of practicing this easy to understand modeling software N In this case , The program is written very smoothly IT Professional classmates , Use Python The prediction model built by the open source package , Accuracy is just not up to it . In fact, Yiming modeling also uses Python Open source package , We all use the same set of basic things . however , Only understand IT Students often do not seriously and systematically learn the principles of these model algorithms , I don't know how to preprocess the data 、 What is the situation that should be corrected 、 What rules should be used when filling in missing values ; I don't know which of these dozens of algorithms has its own characteristics 、 Suitable for what scene 、 What parameters should be filled in . We should study the principle of artificial intelligence algorithm carefully and systematically , It is roughly equivalent to reading half a doctor majoring in statistics , A few months' crash course is out of the question .

However, the experience of statistical experts for decades has been solidified in Yi Ming's modeling , This is only good at IT Of course, entry-level data mining players can't do it . and , Not only is the accuracy better , The work efficiency is also very poor , This thing can make a model in three minutes , It takes a few days to do it manually ; More to the point , This software is free , You can work day and night , How much does a data scientist get paid ？ What do you think the future bosses will think in the face of this situation ？

result , A general data scientist who has roughly learned about machine learning , It is not as effective as those who have industry experience . Rich business experience can also prepare data better , It can also make the model more accurate . Automatic modeling software can only solve the technical problems of artificial intelligence , There is no way to automatically discover business knowledge , Therefore, even if there is automation software, good business knowledge is required to build a good model .

Future workplace , Probably not the world's data scientists , But the automatic artificial intelligence software and industry business experts all over the world .

For example , It's like doing laboratory tests in the hospital now . before , It takes a fairly skilled person to get , Only high-level hospitals can afford it , These people are also very popular ; Later on , Then came the automatic instruments , Everyone can do it , Even a hospital can do it .

---

That is to say , Do you feel a lot of crisis ？ Is the profession of data scientist completely cool ？

Of course not .

Automation like Yiming modeling AI Software also needs people to do it , It needs excellent data scientists to do it . and , Software is not that smart , There are always uncertain situations , At this time, it still needs data scientists who are proficient in the principles of algorithms to solve . however , Mastering the principle of algorithm is the premise , otherwise , You can't do anything with software .

It's just , Of course, it is very difficult to be such a person , And the demand for this kind of person will not be very large , Most routine data science transactions will be replaced by automated software .

---

The profession of data scientist is promising , But only I worked hard N New year's window Top master of . It came out of a street training class or a crash course in a university , Then save it .

---

Yiming intelligent modeling materials

• Product introduction official website
• Modeling product download