Alexander Polyakov
January 22, 2020

2 hours 10 minutes

Design safe AI/ML options

Extra Info

  • Design safe AI answer architectures to cowl all elements of AI safety from mannequin to surroundings
  • Create a high-level risk mannequin for AI options and select the suitable priorities towards numerous threats
  • Design particular safety exams for picture recognition programs
  • Check any AI system towards the newest assaults with the assistance of straightforward instruments
  • Study a very powerful metrics to match numerous assaults and defences
  • Deploy the suitable defence strategies to guard AI programs towards assaults by evaluating their effectivity
  • Safe your AI programs with the assistance of sensible open-source instruments
About Synthetic Intelligence (AI) is actually consuming software program as an increasing number of options turn out to be ML-based. Sadly, these programs even have vulnerabilities; however, in comparison with software program safety, few individuals are actually educated about this space. If it’s unattainable to safe AI towards cyberattacks, there will likely be no AI-based applied sciences, corresponding to self-driving automobiles, and yet one more “AI winter” will quickly be on us.

This course is nearly definitely the primary public, on-line, hands-on introduction to the long run views of cybersecurity and adopts a transparent and easy-to-follow strategy. On this course, you’ll study high-level dangers concentrating on AI/ML programs. You’ll design particular safety exams for picture recognition programs and grasp strategies to check towards assaults. You’ll then study numerous classes of adversarial assaults and the way to decide on the suitable protection technique.

By the top of this course, you may be acquainted with numerous assaults and, extra importantly, with the steps you could take to safe your AI and machine studying programs successfully. For this course, sensible expertise with Python, machine studying, and deep studying frameworks is assumed, together with some fundamental math expertise.

All of the code and supporting information for this course can be found on GitHub at:

  • Achieve sensible expertise with numerous open-source instruments corresponding to ART (Adversarial Robustness Toolkit) and DeepSec, developed to check machine studying algorithms for safety
  • Study to design safe AI options relying on dangers which might be typical to your software with the assistance of a novel strategy
  • Perceive the assaults and completely different approaches for securing numerous AI/ML programs
Course Size 2 hours 10 minutes
ISBN 9781838826451
Date Of Publication 22 Jan 2020

Dimension: 772MB

The put up [PacktPub] Securing Your AI and Machine Learning Systems [Video] appeared first on FreeCoursesOnline.Me.