In 2017, the first Sino US data science comparison report, Python was ranked first in popularity, and the median annual salary of US data workers was up to $110 thousand.
The latest news, Kaggle recently in the field of machine learning and data science were investigated in depth of the whole industry, the survey received a total of more than 16000 replies, the respondents included what is the most popular programming language, what is the average age of data scientists in different countries, the average annual salary of different countries is how much.
However, because China's data collection is not comprehensive, and the US data are also not enough, so, so,The following data are for reference only. I hope Kaggle can make the data more thorough and more thorough next time.
The following is the data collation of AI technology base, and from the perspective of Chinese and American data science and machine learning comparison.
Survey and comparison of Chinese and American data workers
In the world, the average age of the survey is about 30 years old, and of course, there is a change in the value between countries.The following is the age comparison of the respondents in China and the United States:
In China, the median age of the machine learning practitioners is 25 years, and the practitioners are concentrated at the age of 20-30. This may reflect the general distribution of Chinese practitioners, but given the amount of data that Kaggle has made, the details are still questionable.
In the United States, the median age of machine learning practitioners is 32 years, with the largest number of age 20-30. But what is surprising is that we see a cow at the age of 100 in the chart, and several children close to the age of 0. We don't know the details of data cleaning in Kaggle yet, but if these big fruits exist, please contact AI technology camp. We are very interested in your existence.
Comparison of employment status between China and the United States
The total number of workers in China is 53%, and the United States is up to 70.9%
Chinese and American data science specific position comparison map
The field of data science can cover a lot of works, including machine learning engineers, data analysts, data scientists, software developers, data mining workers, etc. The following is a comparison between China and the United States in the field of data science:
Globally, the median annual salary of data scientists is $55441. In China, the median annual salary of data scientists is $29835. The United States is up to $110000
ChinaFull time salary
Full time annual salary in the United States
The highest degree of Education
Generally speaking, the most common academic degree of data science practitioners is master's degree, but generally speaking, a Ph.D. degree can get a high salary ($150K $200K and $200k+).
As far as China is concerned, the master's degree is 40.5%, the doctor is only 11.2%, the number of the bachelor's degree is 39.5%, and the number of master is equal.
In the United States, the master's degree is only 44.5%, the doctorate is up to 20.7%, and the undergraduate practitioners account for 26.5%.
In general, the doctorate in the United States is up to 20.7%, which is two times closer to China than in China (China is 11.2%).
How do data scientists work in the end?
What kind of methods do you use in your work?
Logistic regression is the most commonly used method of data science, in addition to the military and national security fields. In the field of military and defense security, neural networks use more land.
Overall national data
Is the most used tool language in data work?
In general, Python is the most used language for data workers. At the same time, the data researchers are also very loyal to the R language.
Overall national data
What type of data do you use in your work?
Relational data markets are the most commonly used data types. However, text and images are more popular in academic researchers and in the field of national defense security.
Overall national data
What kind of code sharing and hosting are used in the work?
Most data workers use Git to share code. However, large company workers prefer to keep the code locally and share the code with mail. Start-ups use faster cloud sharing.
Overall national data
What kind of obstacles do you encounter in your work?
The dirty data (Dirty Data) is the biggest obstacle. Machines have a focus, but the ability to understand different algorithms is also a big obstacle to data workers. The lack of effective management and financial support is facing two big data workers in difficulties.
How do new data science newcomers emerge in the industry?
According to your experience, which language do you recommend to the new data science newcomer?
This varies from person to person. In the largest language of the two largest use of Python and R, most people feel that Python is more worthy of being recommended.
Where do you get the learning resources of data science?
Data science is a very fast changing field, and the people in the industry need to constantly update their knowledge system to keep a certain position in the industry and not be eliminated by the times. Stack Overflow Q&A, Conferences, and Podcasts are the learning platforms that practitioners have often used. When issuing new software, be sure to remember to read the official use guide and recommend to YouTube to watch the use of video.
Where do we get the open data set?
Without data, there is no data science! When it comes to some data science skills, it is important to know how to find clean open source data sets and projects for practice. More and more people are starting to use our data set aggregator (https://www.kaggle.com/datasets).
By what channel do you get a job?
According to the experience of people in the field of data science, these methods may be more efficient than sending resumes on company website and recruitment website, for example, by establishing their relationship network in this industry.
The above comes from the kaggle website. Because the text is multidimensional to a number of countries, if you want to see the full picture of the industry, please click:Https://www.kaggle.com/surveys/2017
The one hundred day, the Artificial Intelligence Engineer's learning plan -- the whole battle case, from machine learning principle to recommender system, from deep learning entry to image semantic segmentation and poem writing robot, to the four industrial level actual projects on the exclusive GPU cloud platform. A perfect master of the skills of artificial intelligence engineers within 100 days.