Machine Brains Start Learning Unsupervised: Elastic Acquires Prelert

Author

Adrian Bridgwater

September 30, 2016

As we know, planet Earth is now under the control of intelligent machines and, in this new ‘rise of the robots’ dystopia, we have seen machines start to become ‘self-aware’ sentient, emotional and even a little moody from time to time. Okay so that hasn’t happened yet has it? But machine learning and artificial intelligence is on the rise… and so-called ‘behavioral analytics’ technologies are expanding, proliferating and positively diverging all the time.

What happens next in this market is the same techno-corporate process that happens throughout the rest of the industry i.e. firms start to acquire, merge, absorb (and occasionally divest) each other.

Logging business

In this regard then, Elasticsearch (a search logging analytics firm) has acquired Prelert (a behavioral analytics technology firm). As Prelert’s DNA is absorbed into the Elastic software stack, customers will be able to use the resulting analytics set to solve complex use cases in cybersecurity, fraud detection and IT operations analytics and so on.

What makes Prelart (arguably) interesting is that it specializes in unsupervised machine learning’ to deliver behavioral analytics i.e. the type of machine intelligence where computers really start to do interesting things.

Unsupervised learning means no correct answers

According to MachineLearningMastery, “The goal for unsupervised learning is to model the underlying structure or distribution in the data in order to learn more about the data. These [models] are called unsupervised learning because unlike supervised learning above there are no correct answers and there is no teacher. Algorithms are left to their own devises to discover and present the interesting structure in the data.”

Elastic says that as firms become more sophisticated in their use of big data tools, they come across questions which cannot be answered using simple search, so first Elastic has added Graph functionality, now behavioral analytics. Firms can create applications based on live data analysis that would otherwise require an ‘unimaginable number’ of people watching a screen says Elastic.

Logs from firewalls, servers etc, can be used to alert teams instantly as an attack starts. Prelert claims its technology can work out the precursor signs of past problems and alert sysadmins before they happen.

Prelert was founded in 2008 to create technology that automates the discovery of anomalies in large, complex datasets, predicts actions and outcomes. It provides enterprises and their end users with a consumable application that doesn’t require them to perform data science.

“Using unsupervised machine learning techniques applied to a customer’s historical and real-time continuous data, Prelert’s predictive models perform behavioral analytics to understand the probability of failures and events occurring with built-in alerting and notifications for end users to explain ‘why’ something has happened and ‘what’ to do with that information,” explains the firm.

The next machine brain

This is not artificial intelligence as such, this is rather more a case of machine-learning in a slightly more purist sense. The applications could be wide-ranging and security intelligence is certainly right up there as a very suitable application of this type of intelligence.

The next stage, as is intimated and suggested in part of this announcement, is that machine intelligence power will start to move towards users who don’t necessarily hold a degree in data science.

Yes, overall, your computer (potentially) just got more powerful.

 

This article was written by Adrian Bridgwater from Forbes and was legally licensed through the NewsCred publisher network.

Comment this article

Great ! Thanks for your subscription !

You will soon receive the first Content Loop Newsletter