Online classrooms make it easy for people to become data scientists

Many companies are beginning to pay attention to large data and data analysis, but the talent is difficult to find Ah! In fact, it's easy to train a new programmer to be a capable data scientist, just a few cloud servers, and then learn a few weeks of machine learning on the Internet with a data specialist.

The most famous data scientist training case is the "Enterprise Prediction Solution Platform" Kaggle, the latest award winner, Carter S. The Kaggle user has developed an "overkill" analysis tool to predict risks in the insurance industry using a simple but efficient approach.

This is a surprising tool, and Carter found a good job by using the content he had learned in the online classroom for risk forecasting in the insurance industry. He had studied natural language processing and social network analysis before, so big data analysis was hard to do. But how do you make big data scientists who are just out of college and have no experience? Online Classroom crash can!

Luis Tandalla, after learning some free courses in Coursera and other online classes, took advantage of the knowledge he had learned to win a prize in Kaggle's competition, and his work helped teachers to correct simple questions and score them. And he had no idea what artificial intelligence and machine learning were.

Luis Tandalla said that to do data scientists, first of all have to learn the passion. So he took natural language processing and probabilistic template courses on the Coursera, and then he practiced his point of view while studying Kaggle. He will graduate next year with a bachelor's degree in mechanical engineering, rather than a computer science major imaginable. He said he wanted to create a predictive software service company after graduating.

Tandalla may not be the only example. Most of the winners on the Kaggle studied machine-learning courses on Coursera. Singaporean Xavier Conort last year determined to transform data scientists, who have spent only a year in the online classroom to become one of the top data scientists in Kaggle.

Data analysis Guides

Andrew Ng, Stanford University professor and co-founder of Coursera. It is not just a coincidence that the machine semester course he teaches on Coursera is the highest degree of completion in all online elective courses. If you want to go with the big data step, become a data scientist, in the country does not have the advantageous resources, may consider Coursera, Udacity, edx and so on the network classroom, free to carry on the study. In China, many data mining companies should be very fond of such courses, can save a large amount of training costs.

To become a data scientist, it is important to understand algebra and probability first, and the prerequisites include a basic understanding of programming, Ng said.

"Machine learning is becoming one of the most sought-after technologies in Silicon Valley," he said. "Many business personnel officials say that because companies are currently in urgent need of such data analysis talent, so long as an employee can finish the online course on time can significantly improve his salary and career prospects."

Why are such online courses so popular? Why can data analysis change the world?

Ng believes that these online courses are so popular because it turns existing and mature theories into technologies that can be applied, and that students who have learned the prerequisites have the opportunity to perform, rather than just programming and writing programs and writing applications. In addition, students can adjust the pace of learning according to their learning ability, the information on the forum can also help them to complete the course.

Ng said he would not have been able to speak such an excellent machine-learning class without having the privilege of mixing with the world's smartest computer experts in Silicon Valley. In his course, he rarely talked about algorithms, or how to apply machine learning to practice. He thinks it is more important to learn to apply than to learn knowledge. This is like learning programming and learning programming language, one is practice, one is only theoretical knowledge.

Study hard, you may be the next Einstein

In fact, becoming a data scientist is not the ultimate goal, even the champions of the Kaggle competition can not be regarded as the focus of their career. Through online courses can also learn more knowledge, so that the ability of people to learn really useful knowledge, to create more results.

"It makes me wonder," Ng said, "Maybe the next Einstein is a little Afghan girl sitting in front of a computer watching instructional video." ”

Article Source: GigaOM

(Responsible editor: The good of the Legacy)