Invincea's future of cybersecurity innovation
Loading...
As Marc Andreeson is credited with saying, "software eats the world", Ghosh points out that new innovations in software, machine learning, and visualization is likely to replace front line "data sifters" in security operations centers (SOCs) with software that is far more effective at finding attacks in data streams. Software and machine learning is well-equipped to deal with massive data sets in distinguishing signal from noise. Skilled cyber security subject matter experts are good at asking the right questions of the data. Together this will transform SOCs from "watching cameras" on data to skilled investigations only into events of interest identified by the software. The over-abundance of data makes machine learning algorithms more effective, which in turn will make human time more targeted at only relevant events of interest.
The talk not only lays out the fundamental challenges and the direction cyber security innovation is likely to follow, but gives tangible examples from Invincea Labs' on DARPA-funded research in natural language queries over distributed agents, and in automatic analysis and visualization of unknown malware.