To become a hot data scientist, experience is more important than education.

Data scientists are rapidly emerging as the tech giants of the 21st century, in part because of the precision predictions of the American election's shepherd, Nate Silver, and Debord Desta (Paul DePodesta), a baseball-tactical revolution based on data analysis, They let the math experts who are immersed in numbers evolve from nerds to stars, but over the years, people have debated how data scientists work with statisticians.

As the field of data science grows, many companies are more eager to embrace their own data scientists. However, many people may not expect that these math nerds are not as "educated" as you think.

Many of the best in the data field do not have the math or science training that many expect, Shepherd and Debord Desta are just Bachelor of Economics and both have no ph. D., former Facebook data scientist and Data Systems processing company Cloudera co-founder Hammerbach (Jeff Hammerbacher) is also only a Bachelor of mathematics.

"In fact, I think it is often time for Dr Kaggle to think about which algorithms to use, and instead ignore general problems, such as which variables (or features) are important," said Anthony Goldbloom, chief executive of Gaudebru, a huge data science platform. 」

Condido, a data scientist, agrees: "Mathematical knowledge is important, but it is important to understand the study." It is more important to understand the reason for using a certain kind of mathematics than to know mathematics itself. 」

Condido is a master of psychology with no ph. D. in mathematics or physics. But it has grown quite well in the field of data science, and is now zestfinance in the Merrill of the former Google information chief, Douglas Merrill.

Condido says that although the institute's curriculum lays a good foundation for his statistics, it still lags behind the practice in the field of data science. He recommended to participate in the Kaggle-sponsored data Mining contest.

Condido said: "I do not want to belittle the value of doctoral degrees, but do not feel that this is an absolute requirement." 」

Merrill agreed: "We are engaged in data scientists from all walks of life, there are a variety of professional backgrounds, and our team also has no graduate school education ... Because when it comes to data science math is only half of the problem, it's also an art because these people have to have intuition and be able to look at things creatively. 」

Some software companies sell business intelligence applications that can save data mining manpower by using data scientists ' rarity and expensive concepts. Data analysts and business intelligence experts often know what data to analyze, but the methods are more experimental, and they have to find the data mix, figure out what information can be mined from it, and how to dig.

Software on the market can simplify the process of data processing, but data science is not just digesting numbers.

Condido says that regardless of the academic qualifications of data scientists, they are always learning. "It is vital to keep abreast of the latest knowledge in the field, otherwise it will soon fall behind," he said. "Participation in the data mining competition is one of the ways to keep up with the times, always pay attention to opponents than you." 」

(Responsible editor: The good of the Legacy)