Ashrith Barthur, PhD
Security Scientist
Is it Time for Security to Use Artificial
Intelligence ?
Overview
● Analysis vs. Analytics
● Complexity in Cybersecurity
● Biggest Problem of Cybersecurity
● Examples
● Analytics with ML
● Advanced Analytics
Analysis vs. Analytics vs. AI
● Analysis - Study and understand collected data
● Analytics - Data analysis with predictability.
● AI - Analytics with Intelligence
Complexity in Cybersecurity
● Complex Protocols
● Massive Amounts of Data
● Attack Asymmetry
● Distributed Attack Surface
Biggest Problem of CyberSecurity?
Exposure of PII - Personal, Identifiable
Information
Affects:
● Banks and Financial Institutions
● Corporations
● Governments
Examples
● Office of Personnel Management
● Sony
● Home Depot
● Target
Analytics with ML
● Random Forest
● Gradient Boosting Machines
● Deep Learning
● Neural Networks
Advanced Analytics? Why?
● Multiple Sources of Data
● Dynamic contexts for Analysis
● Extremely low latency in prediction.
AI? Why?
● Multiple Vectors of Attacks
● Many 0-day vectors
● Wait time with analysis is a lot
What is H2O Doing?
● Leveraging H2O for data analysis and
analytics
● Building Dynamic Contexts to Analyse data
● Providing Intelligence in sourcing, combining
data.
Thank You
Questions?

Cybersecurity with AI - Ashrith Barthur

  • 1.
    Ashrith Barthur, PhD SecurityScientist Is it Time for Security to Use Artificial Intelligence ?
  • 2.
    Overview ● Analysis vs.Analytics ● Complexity in Cybersecurity ● Biggest Problem of Cybersecurity ● Examples ● Analytics with ML ● Advanced Analytics
  • 3.
    Analysis vs. Analyticsvs. AI ● Analysis - Study and understand collected data ● Analytics - Data analysis with predictability. ● AI - Analytics with Intelligence
  • 4.
    Complexity in Cybersecurity ●Complex Protocols ● Massive Amounts of Data ● Attack Asymmetry ● Distributed Attack Surface
  • 5.
    Biggest Problem ofCyberSecurity? Exposure of PII - Personal, Identifiable Information Affects: ● Banks and Financial Institutions ● Corporations ● Governments
  • 6.
    Examples ● Office ofPersonnel Management ● Sony ● Home Depot ● Target
  • 7.
    Analytics with ML ●Random Forest ● Gradient Boosting Machines ● Deep Learning ● Neural Networks
  • 8.
    Advanced Analytics? Why? ●Multiple Sources of Data ● Dynamic contexts for Analysis ● Extremely low latency in prediction.
  • 9.
    AI? Why? ● MultipleVectors of Attacks ● Many 0-day vectors ● Wait time with analysis is a lot
  • 10.
    What is H2ODoing? ● Leveraging H2O for data analysis and analytics ● Building Dynamic Contexts to Analyse data ● Providing Intelligence in sourcing, combining data.
  • 11.