Machine learning key to building a proactive security response: Splunk
Growing demand for business relevance around security analytics will see machine-learning algorithms playing an increasing role in the large-scale analysis of security logs using big-data analytics tools, the head of analytics firm Splunk’s security business has predicted.
Noting that evolving data architectures have positioned analytics platforms like Splunk as “the nerve centre for enterprise security operations,” senior vice president of security markets Haiyan Song told CSO Australia that the real-time aggregation and processing of a wealth of security information was fast becoming an enabling technology for companies’ data security.
“We’re taking information from all the sensors that a company has in its network,” Song explained, “then bringing it together to syndicate, correlate, and derive all the intelligence that feeds back into their security operations to either operationalise the threat – or even to go one step further, and automate some of the remediation as well.”
The empowerment of security systems to drive defensive measures represents a step forward from conventional security and analytics infrastructure, which generally positions analytics technology as an enabler to drive manual intervention. But with machine-learning techniques becoming increasingly intelligent, Song said, defensive systems would become increasingly proactive – particularly in the wake of the company’s July acquisition of behavioural-analytics specialist Caspida.
“Using a baseline to look for anomalies is super-critical in trying to do early detection, and therefore shortening and minimising the impact,” said Song, whose previous role with HP’s ArcSight security-analytics business have helped her drive the integration of contextual network information into the security response.
“We’re adding more intelligence to help customers really automate the analysis,” she continued. Caspida’s applied-analytics tools utilise data science techniques such as classification, statistical models, Markovian algorithms, inference and grouping models to quickly pick out anomalies and build a security ‘kill chain’ designed to speed remediation by security staff.
Business has been surging in Australia, Song said, as as Australian businesses warm to the promise of cloud-based analytics as a security enabler. Demand has been so buoyant that former ANZ country manager Dan Miller was recently promoted to become the company’s APAC director of cloud – where he will drive growth in the increasingly popular cloud-hosted analytics services market.
“The thing we’ve seen from customers with respect to our cloud offering is the ability to provide hybrid services,” Miller said.
“Since there are always going to be workloads on premises for the foreseeable future, there will be a lot of focus around aligning with customer strategies around their data architectures. Customers understand that there no longer needs to be a tradeoff between managing workloads on premises and losing visibility.”
Australian government and business customers were at the vanguard when it came to embracing cloud-hosted analytics. “Australia is one of the regions that are most aligned with the dynamics in the US,” Song said, noting a series of new Australian government customers that the company was closing “pretty much every month”.
“That’s great momentum,” she said. “Australia is probably more ahead than the US in terms of the willingness and readiness to go to the cloud.”
Interest in cloud-based solutions was tied to customers’ growing desire for solutions, she continued. “Customers are increasingly telling us not that they want a stack of technology, but that they want us to help solve a security problem. And that’s why we’re bringing the data scientists and machine-learning technology to the customer in one black box – so they only have to look at details of the threat rather than trying work out how to find it.”
Want to know more?
Why not become a CSO member and subscribe to CSO’s mailing list.
Get newsletters, updates, events and more right here
Join the CSO newsletter!
Source: Machine learning key to building a proactive security response: Splunk
Via: Google Alert for ML