Big Data + machine learning: Microsoft hopes to regain its dominance

Machine learning (Machine Learning) is a study of how computers simulate or implement human learning behavior in order to acquire new knowledge or skills, and to rearrange existing knowledge structures to continuously improve their performance. It is the core of artificial intelligence, is to make the computer has the basic way of intelligence, its application in all fields of artificial intelligence. Not long ago, the New York Times reported that Microsoft was applying machine learning to the business. Lightspeed, US investment director Jeremy Liew, also introduced the example of "Big Data plus machine learning" to reshape the bank's credit industry.
Application of machine learning technology, software and services are profit points

In the next office release, Excel can combine large amounts of data. For example, you can scan 12 million Twitter messages and then generate a chart that tells you which Oscar nominees are the most talked about. The new Outlook version adds features that evaluate users ' email reading habits and determine which messages users want to read. Microsoft's machine learning software will be able to crawl the company's computer systems, just as the Bing search engine crawls Web pages and links on the Internet.

The explosion of data from sensors, connected devices, and cloud computing centers has created a large data industry. Computers need to find patterns in the mass of daily production. In the long run, Microsoft hopes to be able to use more machine learning techniques in its cloud computing platform Azure, such as leasing data sources and algorithms that allow companies to design their own data prediction engines. Microsoft can eventually charge for software services, rather than just selling software.

Microsoft has something that startups don't have: huge reserves of money--earnings from the end of last quarter, a $67 billion trillion in cash and short-term investment--and the ability to spend 10 or even 20 years on a big project. "Microsoft has too many resources, Windows, ie, Skype, Bing search, etc., and they can do a lot of things," says David Smith, a senior researcher at Gartner. Data analysis will be their next big deal. ”

Fragmentation linkages, "alternative credit" patterns emerge

Some start-ups use massive data mining and algorithms to do some loan business.

Wonga is a start-up for an emerging alternative loan. Sonali De Rycker, an investment agency Accel, is a member of the board of Directors of Wonga, who said: "They use social media and other web tools in large numbers, but these are absolutely unimaginable." And that's where their miracle is. ”

The key to the problem is the algorithm, how it puts your zip code, the color of your car, how much your mortgage is, how it can relate your fragmented stuff. These are the key to Wonga data accumulation and collation of the various pieces of information that it used to be a customer. When Accel entered the Wonga board in 2009, Wonga had 100,000 loan cases. These 100,000 data messages are continuously integrated and categorized into a growing network of information. Each lender has 6000 to 8,000 data.

"You have a lot of data strung into a story. We are willing to pay for the data because we need it. We can use thousands of combinations to determine whether something is right or wrong. ”

Accompanied by a large number of data sources and powerful data analysis tools, it also means that it can borrow at lower interest rates than other payday lenders. Machine learning is based on bad loans, so the more the number of failures, the more you pay tuition, the same your model will be more perfect.

The big data age has more complex business models, but it is only superficial. There will be more innovation in this field over the next few years.

(Responsible editor: Schpeppen)