I asked the pro variously how to become a required data scientist human resources, 'If the skill is not enough, the salary may be reduced'



Among the advanced IT personnel, it is not able to keep up with the rapidly increasing demand, and it is a chronic 'seller market' job profession called ' data scientist '. What is the reality of the data scientist job, which is said to be a drag from companies? So , I asked Mr. Kauchi, who works as an active data scientist in a major human resource persol carrier, and talked about it in various ways, and which course is useful in the world's largest online learning education platform ' Udemy '. I also taught you.

World's Largest Online Learning Platform | Udemy


table of contents
◆ 1: What is “data scientist”?
◆ 2: The reason why the “data scientist” pulls off
◆ 3: There will be a time when salary will decrease if you can not use data science! ?
◆ 4: What is the learning method to become a 'data scientist'?
◆ 5: 5 Udemy data science courses recommended by professionals

Ms. Kaname Shikauchi who has moved from a researcher to a business world and is currently managing the business development of a service called Data Ship, a new business of a major human resource persol carrier, and the data science project of a career change service, Mididas .


◆ 1: What is a 'data scientist'?

GIGAZINE (hereinafter, 'G'):
This time, let us talk about the job of 'data scientist' whose demand is rising rapidly, from the viewpoint of being engaged in data scientist learning support and career support. First of all, please introduce a brief self-introduction of Mr. Shikauchi.

Mr. Gaku Kashina (hereinafter referred to as 'Kanuchi')
I received a Ph.D. in Science from Nara Institute of Science and Technology. At that time, in order to study the human brain, I measured eye movement and measured brain activity called fMRI. In my first job, I taught at the Graduate School of Medicine at Kyoto University and then moved to the Graduate School of Informatics to conduct research. As the last two years as a researcher, I was involved in a national project, and I was doing research such as “open and close the door using EEG”, MRI, fMRI, NIRS, EEG We are also focusing on data accumulation of various brain activity measurement such as EEG.

G:
Did you do research like 'connecting the brain and the computer'?

Shikanai:
That's right. This technology is also called brain computer interface (BCI) or brain machine interface (BMI).

G:
Has your brain research turned into a business dealing with data?

Shikanai:
That's right. In the case of brain waves, 'time series data' comes out. This is analyzed by machine learning etc., and information is taken out. The brainwave data is so noisy that it is difficult to extract information from it, even if there is data. There was technology and experience to solve it, and what I expected on the business side was that it was related to data.

G:
Did you have any opportunity to change from the academic world to the business world?

Shikanai:
I got some offers from the business side.

G:
Do you already feel as a 'data scientist' at that time?

Shikanai:
No. At that time, I did not realize that I was a profession of data scientist.

G:
Three or four years ago, the term data scientist itself was not so common.

Shikanai:
It is not general. Well, I knew that there was a word, but I did not know in detail what kind of occupation it was, and I never thought that I would fall under it (lol)

G:
There are times when researchers don't know that the research they're doing is being evaluated like that in the world ....

Shikanai:
Actually, it is still there. Many doctoral graduate students and researchers are not aware that they have a data science job they can do. So I feel that 'to know' and 'to notice' are quite important.


G:
Is the 'field of vision' of the data scientist extremely large?

Shikanai:
It's wide. Also, I feel that it is not a profession that can be defined by the type of work. I think that is not a general idea, but I think that it is the data scientist's job to do business development after understanding the data.

I am a little disoriented, but there is also the problem of calling. I use the word 'data analyst' or, depending on the industry, call it 'digital marketer' or call it 'machine learning engineer'. The usage of words was different, and when I was in the academic world, I could say 'data scientist' (laughs). This nuance doesn't work on the business side and I was a bit surprised, but it is something that surpasses data scientists. 'What about data scientists is this?' For those who are studying machine learning in the academic world, the level was overwhelmingly low.

I am also not a researcher of machine learning because I am also a researcher of cognitive neuroscience. It is low even if it sees from there. Even if you call it 'data science', the work is only tabulating in Excel, or you clearly misinterpret the analysis results .... Some people are just making graphs if you make a pole argument. So, on the business side, people who are called data scientists saw a situation where the words and the actual situation didn't match about the eyes. As a result, there may not be many people who say that they are data scientists unless their department also names it 'data science'.


G:
Recently, I have come to listen to the word data scientist itself, and I think that many people are grasping as an image, but not clearly. What are the services that are made possible by data scientists' activities so that they can be more easily imagined, and the specific services we use?

Shikanai:
Well, for example, Yahoo! Recommendations when shopping for something is a result of data science. Recently, there are audio speakers with built-in AI. You should also use machine learning. Besides that, ABeam Consulting and other consulting (farm) also use data science. We also use data science to find out how to transfer, deploy, and adopt personnel information.

G:
Basically, everything related to machine learning can be considered to be related to data scientists?

Shikanai:
Data scientists are involved. Data science does not necessarily use machine learning because it can contribute to sales increase even by using a general statistical analysis method called “linear regression analysis”.

G:
Simply put, is it that a person who can create value out of information is a true data scientist?

Shikanai:
That's right. It is important to generate value from data. Data scientists often tend to focus on data analysis and evaluation, but sometimes they start with 'data utilization design' and 'data acquisition and analysis design'.


Shikanai:
For example, in the case of Bridgestone's, I attached a sensor to a tire seven or eight years ago. In the first place, while the tire is spinning at high speed, it is difficult to take data while flying data wirelessly, but it is now possible to do business because of the data acquisition. Data from tire sensors can tell you if the road surface is dry or frozen. If the expressway is frozen, snowmelt agents and so on are used, but until now the person visually did the judgment, and it was said that people actually touched the road surface if they did not know visually. But isn't it dangerous to do that on the freeway? It seems that that is solved by taking information from the tire sensor.

G:
I see. It seems that the viewpoints tend to gather at the point of analysis and evaluation, but it is impossible to finally get something good without getting involved with the data of the previous stage.

Shikanai:
I can not do good things. Ideally.


◆ 2: The reason why “data scientist” pulls off

G:
The demand for data scientists is growing rapidly, is that basically all areas are required to make good use of data?

Shikanai:
We sought. I think that is connected to what is said that 'data scientists are overwhelmingly lacking'.

G:
'I have extremely high demand of data scientist, career ratio of job offers to six times' that data and, (PDF file) of 'shortage of 48,000 people at the tip IT human resources 2020' data there is. How many 'leading IT human resources' currently have?


Shikanai:
There are around 110,000 in 2016 (PDF file) data issued by Ministry of Economy, Trade and Industry. On the other hand, about 30,000 people are short. What this 'lack' means is 'If you have a data scientist here, you can do new business.'

G:
It looks like an opportunity loss.

Shikanai:
Yes, there is a loss of opportunity.

G:
Is the situation where supply has not kept pace with the demand of data scientists?

Shikanai:
Yes. As the scope of data utilization is expanding, data scientists can contribute to profits and sales, but there is no human resource. Of course the number is increasing, but there is more demand than that. Until now, it was in the direction of using data acquired on the Web for other purposes. However, we are planning to work more actively and actively, and like Bridgestone's earlier, it is in the stage of trying to get data from the beginning and to get usable data.


G:
Do you feel that companies outsource data scientists at this time?

Shikanai:
No, there are an increasing number of companies making in-house. In general, we will work with vendor companies to create cutting edge products internally so that we can do it ourselves. In data science, too, there are an increasing number of companies that are producing in-house. So it is not enough. Well, I can't make it in time because I can pull it.

G:
I see. It's a story that companies want to have data scientists in-house as well.

Shikanai:
That's right. User companies will be able to analyze their own data properly. This is a sale, but I think data scientists are 'multiplying people.' I think that the value of such human resources will increase in the future. Of course, as technology is refined in the future, it will become increasingly workers. I think someday data scientists will either become blue workers or people's professions. However, at this point in time, it is a human resource that works in 'multiplication.' It is important to choose a good data scientist, as it is a human resource that can be multiplied by productivity.

G:
It is completely different only by the difference of 1.2 times and 1.3 times and 0.1 in multiplication.

Shikanai:
It's totally different. In addition, even if there is only one person who has double productivity, it will only divide 100 departments into 101 people. Only 1% change. If you have 10 billion yen in the original sales, you can get 12 billion yen, or 20%, even for 1.2 times of human resources. Furthermore, if it is 1.3 times, it will be 13 billion yen. Moreover, although there is only 0.1 difference between 1.2 and 1.3, 1 billion yen is also different.

G:
Does the quality of data scientists change so much?

Shikanai:
I think it will change tremendously. There was a development example created in the classical statistical way, but if the development is carried out using machine learning, the development period is 1/3, the number of developers is 40%, and the cost is 1/20. It was achieved in moderation. Good people, good team building can be a good example of low cost and short duration. When you multiply various efficiencies and values, they may differ by more than 100 times.

G:
The magnitude of such a 100-fold difference depends on the quality of the data scientist.

Shikanai:
It depends on the quality. The previous example is a successful example, but I think it will change about 10 times even if it is not so.


◆ 3: There will be a time when salary will decrease if you can not use data science! ?

G:
If the demand for data science is growing, much more than a high quality data scientist, then companies want good people. In a situation where there is an acquisition competition for such personnel, is that, isn't the data scientist really getting preferential treatment in terms of compensation?

Shikanai:
Still, Japan is the current state of the future. That's because, depending on the industry, there are no data scientists in the first place, so I don't know what value they produce.

G:
That's right, unless you realize that you can use data to make a profit in the first place, you can not easily prepare it.

Shikanai:
As a company, you can not create a personnel system for people who are not. You have not been able to 'value' the data scientists correctly. From my point of view, I think that if you make good use of it, it will work in any industry. You need data scientists in every industry. Of course there are shades. There is a saying, 'It's not going to work well at this time.' If it is properly identified, high money will be paid, but I think in Japan it has not come yet.

G:
For example, do you use it better in America?

Shikanai:
I agree. I think there are many mistakes in the United States, but I think it is easier to take risks because there are differences in employment practices. In Japan, especially, large companies often take data scientists in general occupations. Because of that, it is difficult to become a professional, and the personnel system is the same. It is not possible to raise salaries because it is not possible for data scientists alone to have a special salary system.

G:
I see. However, IT technologies, trends, and business styles often come from the United States or are introduced to Japan a little later. Japan also seems to be changing with regard to the employment system, so it is clear that Japan's human resource utilization will be successful from now on based on the United States as a good role model?

Shikanai:
I think that is no mistake. Then, IT companies such as Yahoo Japan Corporation have already hired data scientists as professionals. In addition to good salaries, there are also follow-ups for continuing to study as a member of society. Other IT companies seem to back up travel expenses to international conferences and learn in working hours and create opportunities for skill improvement. As such, some companies have already enhanced data scientist treatment and learning support.

In the future, data science will be a common technology. It's not that office workers can not use PCs right now. In that sense, it may be better to think that if you do not have the skills, you will lower your salary, than if you have the data science skills to improve your salary and treatment . It may be said that you should learn for that too.

G:
It is a little scary story ... (laughs)

Shikanai:
Of course, now we should seek high rewards. As an employee of 'contract without fixed period' in October, I quit my career career. Since November, we have an individual contract with Persol Carrier's new business, Data Ship. I signed a professional contract as a business developer who creates data business. We also contracted with Mididas, who is a carrier of the parsol, to promote data science projects. As a professional data scientist, we are moving into the next phase. In addition to simply receiving a reward based on the contract term, we have set an option to receive additional rewards if it is successful. I'm looking for a new way of working.

G:
If you are working as a data scientist, what kind of people do you have? Are you originally programmers or SE people?

Shikanai:
I think it is not limited to programmers and SE. That's diverse. Starting with tabulation and data analysis in Excel and Access in marketing, the need for database maintenance has emerged, and there are also examples of those who have built system infrastructures in a stable manner. Although it is special, there is also an example where the former job of a woman who analyzes data of human resources is a nail list. She seems to have been a literary system, but when I listened to it often, I was a person with some mathematical background such as symbol manipulation. This is also true for those who have studied psychology in the sense that there is a cultural background. Psychology is usually found in the Faculty of Literature, but you do the basics of statistical science. People tend to be found in the literary system of Kotekote, but some of the people who are doing research are able to analyze statistics. I want them to realize that they can be active as a data scientist.

G:
There is probably nothing that can not collect data, so it can be used in all fields, but it is likely that you will not realize that you can use it.

Shikanai:
Speaking of the way people who have studied psychology go, as an expert, I think that there are many who become counselors and go on the way of education. But isn't it also interesting to have a career that does people analytics (human resource data analysis) with “human resources” that opens up a little bit of perspective?


G:
Certainly, when it comes to human resources, there is a section that has tended to make a fluffy decision on things that have not been made into data or senses or relationships until now. However, it is an area where there is an opportunity like that if you can get data and use it, you will get better matching.

Shikanai:
It is a chance that the data has not entered so far . The manufacturing industry mentioned earlier is one of them. Retailers also use POS data, but not enough. So is the restaurant reservation system. Mr. Toreta is not only analyzing visitors but also analyzing and automating seating arrangements. There is a table for four at the restaurant, but 'Can I put two here or four?' I think it's interesting to analyze data, see tasks and improve them.

G:
While there are many companies that do not use data science at all, there are companies that are already using it to grow more actively.

Shikanai:
I think this is just the turning point of the industrial structure. Web services that have been in existence since the 2000s are now in a row from 10 to 15 years. The IT industry is trying, for example, to create Google Glass by Google, or to move into a real world rather than a virtual world. Amazon is also making a real store. Furthermore, I am moving to the world of IoT, so I want manufacturers (manufacturers) to make good use of data science. Informatics and data science should be able to create synergetic effects when things like manufacturing can be made.

G:
Does this mean that manufacturing is an area where data is not yet used?

Shikanai:
It is not used. The know-how of the manufacturing industry can not easily be copied, so it should be a leader in that respect. But, conversely, there was a reason for not having introduced informatics yet. The IT industry didn't have anything, so in a sense, we have been incorporating information science and data science more and more. In terms of manufacturing, for example, in the case of automobiles, there are many mechanical and chemical personnel, but relatively few information systems.

G:
In the world of self-driving cars, it seems that the balance will be worse.

Shikanai:
It is often not so-called data driven. Although it is a matter of degree, although numerical simulation is carried out using models and knowledge, I hope that if the design of the issue is well done, more results can be produced from the data.

There is voice recognition called Siri of the iPhone. Ten to fifteen years ago, Japan's speech recognition technology would have been in the top five, but I remember that when Siri first came out, it was not as practical yet. You can use it practically enough now, right? I think that there are a lot of parts that have been improved since the data is stored. There are innovators who release the service for the time being and use the new service in an entertaining way, and the data is stored accordingly. There is a part where technological innovation advances if a system for accumulating data is built, so we will implement it for the time being. I feel like I won't work if I don't think while running.


G:
The more IoT can get information and data, the more data is growing and the scope of use is likely to expand.


Shikanai:
I agree. In that sense, I would like engineers and data scientists to do a 'design for data utilization' stage before data analysis, that is, how to make money in the business. I'm glad that there are more people interested in it. For example, Rearit Co., Ltd., who is responsible for the data science of Don Quijote, who is known as a comprehensive discount store in the retail industry, is working on new data acquisition. Otherwise, Japan will be in a critical situation in the future.

G:
Are you inferior in international competitiveness?

Shikanai:
That's right. For example, many of the founders of so-called 'GAFA' such as Google and Facebook are taking Ph.D in physics, or they are from science. In Japan, there are not many business people who are from science and technology. In the manufacturing industry, some executives have doctoral degrees, but they are not experts in data utilization. I think it would be crazy if people with high expertise in science and technology make business properly and take responsibility for money. The more profitable the service is, the more data (collection and analysis) it is, so the profit and the data are inseparable . Such data will bring about technological innovation and synergy effects. While I can be a human being and a data scientist, I also want science and technology training at graduate school etc. to be involved in business development and management, as well as data scientists.


◆ 4: What is the learning method to become a 'data scientist'?

G:
So far, I have heard about the data scientist's work and the current situation, but please tell us about the issue of 'How can I become a data scientist?'

Shikanai:
Currently, Data Ship regularly holds off-line events and offers data scientist talks and information exchange opportunities for various types of business types. It also offers online learning and Udemy recommends it as one of them.

Udemy is an online learning platform operated by US company Udemy, Inc., used by 24 million people worldwide


Shikanai:
Udemy is a video, so it's good to be seen everywhere, such as on a train. In Udemy, it's easy to look at it. Along with other studies, it will be Ali to use for review, and the ease of animation is overwhelmingly large. The same is true for comparisons with books. It's pretty hard to take a book and walk, but with Udemy you can see it on a smartphone that many people carry with you.

G:
If you specifically want to become a data scientist, learn in Udemy specifically, how should we learn?

Shikanai:
Learning programming is also the first step.

G:
As a first step to data scientists, programming is a major premise.

Shikanai:
I think you should learn. Actually, Data Ship accepts an internship as a business, but at first I am working in Excel. After two weeks, I notice the limit. It's bothersome to do the same work over and over (wry smile). CSV files, even if they fit into one, the data does not have to be large. It will be a success if you think that it is troublesome.

Languages are centered on SQL, R, and Python to study data science. SQL is a language used to retrieve data from databases, so using it makes data analysis much easier. On the other hand, if you can not do SQL, you have to ask a database engineer, which takes time. It seems better to remember about SQL so that it can be turned early.

G:
How should I use R and Python as a language of analysis?

Shikanai:
Python is used in the engineering system, and it is popular as a language that allows everything from data analysis to system implementation to be consistent. However, I also choose R because I think that R is also a hand and there are many people who are familiar with it. Since both R and Python have statistical 'libraries', I think that either is fine.


G:
If it is Python if it is possible to the system development ahead of analysis, it will go smoothly.

Shikanai:
I agree. And it seems that Python engineers tend to have high annual income. I think that is one of the attractiveness.

G:
What do you need to learn if you are familiar with programming and want to learn data science yourself as a person who is actually working, or still aim to be a high quality data scientist?

Shikanai:
To 'visualize' to learn about data. Keeping spreadsheets out of numbers, or, more precisely, understanding the distribution of data. The next step is to properly design data science from purpose design, such as 'What kind of decision do you use for graphs in the first place?' Don't fall in love with making an analysis algorithm and satisfying you.

In that respect, I think that what SIGNATE (formerly Opt Works) made in the course of Udemy is good. There are specific examples coming up there. The part 'Do you make it for what purpose?' I think it's very good to be able to learn while conscious of 'how to use data'. That is probably the big difference between data science and engineering. If you try engineering, you can see if it's a movement you're assuming. It is hard to test for unexpected movements.

However, in data science, if you make an algorithm and predict it, the result will come out for the time being. However, there are many cases in which you make a “wrong evaluation correct”. You should also learn to assess whether the results are really correct. This is the part called 'validation'.

For example, there are many announcements such as '90% achieved in accuracy rate!', But it is often wrong to evaluate '90%' as 'good'. For example, in the human resource industry, if you aim to 'predict whether the person will be successful or not' based on the data that '3 people will be hired if 100 people are received', you can actually think of a method that hits 97%. Shane. It is an algorithm that says 'not adopted'. In this case, most people will not be hired, so if you predict that you will not be hired, 97 out of 100 will be hit.

G:
But that doesn't mean it, right?

Shikanai:
It doesn't make sense. As it is adopted or not, it seems to be 50%, half a half, but 97% will definitely hit if one knowledge is put. But this would be the same as not predicting. The place to make this 97% to 98%, 99% is really a game place. If you just say 'I hit 97%', that doesn't mean anything. There are a lot of cases announced in press release without noticing it.


G:
It is high as a numerical value, but it does not reflect the substance.

Shikanai:
It has no meaning. It is useless if you do not evaluate in consideration of it. Because engineers tend to overlook that point of view, I think that validation data can be a good data scientist.

G:
Can you learn such things?

Shikanai:
of course. At first I did not know at the beginning of my twenties. There is also a data science course that mentions that properly.

G:
The background to becoming a data scientist is diverse. Under the present circumstances, is it possible to become a high-quality data scientist starting with a self-study from scratch?

Shikanai:
Depending on the level you are aiming for, it may be difficult for you to study alone. It is difficult even if the company's OJT (On the Job Training) alone. In a company, it is called 'person with 10 years of experience'. This is measuring skills with work experience. However, people with 10 years of data science experience at a company are not quite there.

At universities and research institutes, young graduate students are conducting research projects involving data at the forefront of the world. In the world of data science, young people are full of skilled people. If you have that experience, you can use it without much work experience or work experience, so I would also recommend it.

G:
In short, does that mean you can become an immediate force without any experience?

Shikanai:
Yes. Experience is more than just 'job' experience. Conversely, if you left the university 20 years ago, you should not be able to catch up with the cutting edge if you do not have the habit of reading papers after entering a company, so it should be a disadvantage in that sense is. I think it is necessary for Udemy to review the basics one more time.

G:
If you are constantly learning and not trying to catch up on your own, you will be left in no time.

Shikanai:
Regardless of data science, probably in the current era technology development is overwhelmingly fast. I used to go back to the business side after about 20 years when I used to work at university, but my sense is that data science has come down in about 10 years. doing. After all, there is also a lot of data stored on the business side. It's not only a university, but it's also at the cutting edge of business in some research. I think data science is a situation where academics and business are stuck.

G:
So it's important how quickly you bring academic technology to your business.

Shikanai:
It is important. Graduate students are also important in terms of human resources who handle technology. As a graduate student, it is hard for companies to do 'research involving data analysis for 2 to 5 years'.

If you were researching at graduate school two years ago, even if you were not involved in data science work now, I would like you to see Udemy's machine learning course. You should be educated to develop the basics of data science, so you should feel, 'Well, you can do it for yourself?' If you're out of the front line, you'll start reviewing and working on Udemy's video course and start over. Then you may be able to start another career.

◆ 5: 5 Udemy data science courses recommended by professionals

G:
Finally, Mr. Shikauchi listed five recommended courses in advance for those who are going to become data scientists from now on. Please briefly explain the merits of these five courses.

◆ 1: 【Data analysis starting from scratch】 Introduction to Python data science studied in business case


Shikanai:
'[Data analysis starting from scratch: introduction to Python data science to learn in the business case' can be recommended most personally. It is also good to enter from the point of creating the program environment. There is a lot of installation needed to create a program development environment, but it's surprisingly annoying and annoying. People who have never run a program may be the most successful.

In addition, because it is composed of two types of business cases: 'sales forecast for lunch' and 'targeting for customers in the bank', I think it may be interesting for people who are not programmers or engineers.


It is content that can proceed with data analysis if you are interested in a familiar case, and since it is included up to validation, all processes after acquiring data are included. You can experience the data science work in a company.


Other than that, it is also good that the beginners are focused on the minimum necessary to work on. Too much information can cause confusion. The most used algorithm of regression analysis and classification is introduced. In many companies, about 95% would be enough. Now that deep learning is hot, I think that all engineers bite there, but I think that few companies have data that can apply deep learning. It is fine if it is a condition that data is really large in an IT company. It is good that it is explained that the very first such as linear regression etc. is used first.

◆ 2: Introduction to SQL and data analysis for the first time-SQL beginners course for business people to utilize database data in the field


◆ 3: Super practice! 'Business data analysis to learn in R' course


Shikanai:
The “Introduction to SQL / Data Analysis for the First Time” and the “Super-Practical! Business Data Analysis Course Learned with R” are also composed of basic and beginner-friendly curriculums with “satisfying composition”. Too many novices can get too much information because they are too many, and they can not be used well if they are not enough.

For the above basic course, I chose the following two because I can learn advanced content.

◆ 4: 【Learning with TensorFlow · Keras · Python 3】 Introduction to time series data processing (RNN / LSTM, Word 2 Vec)


◆ 5: Basic Deep Learning Courses Fundamentals of Deep Learning Learned in Chainer and Python from Zero


Shikanai:
'[Learning with TensorFlow · Keras · Python3] Introductory to time series data processing (RNN / LSTM, Word2Vec)', specifically, to learn how to use a convenient library to create advanced algorithms, Google You can learn 'TensorFlow' developed by. You will also learn how to use Keras, which is useful for creating models of neural networks. Using two more, you can learn word2vec, RNN and LSTM algorithms with concrete examples such as machine translation, emotion analysis and stock price forecast. Development environment construction such as installation of the library is described in both Windows and macOS and is very kind.

'Everyone's Deep Learning Course' uses the library Chainer developed by PFN, an AI start-up that Japan boasts. The explanation of Python is also available from 'Hello world'. If you are an engineer, you may start with this course. You will also learn about mathematical concepts related to neural networks, such as backpropagation, activation / loss functions, and gradient descent methods.


Shikanai:
Data science is the most different from engineering in the evaluation of prediction accuracy, so in addition to learning in Udemy, concepts and methods for evaluating prediction accuracy, such as temporary testing, cross validation, information criteria, etc. I would recommend learning.

G:
I see. Thank you for all the valuable stories today.

◆ Udemy Fall Sale (Black Friday & Cyber Monday Sale)
Black Friday & Cyber Monday Sale is being held on the online learning platform 'Udemy' that came out in the interview. You can take up to 95% off until November 28, 2018. For those who want to acquire new knowledge and skills, or who want to improve their skills, this is a great opportunity.

In addition, one review of recommended Udemy carefully selected 10 courses can be confirmed by the following articles.

【Latest Lectures】 Udemy Fall Large Sale where technology such as AWS ・ AI ・ Blockchain can be learned by animation 【Up to 95% OFF until 11/28 17】-GIGAZINE


in Coverage,   Interview,   Software,   Advertisement, Posted by darkhorse_log