Artwork

Το περιεχόμενο παρέχεται από το Skyflow. Όλο το περιεχόμενο podcast, συμπεριλαμβανομένων των επεισοδίων, των γραφικών και των περιγραφών podcast, μεταφορτώνεται και παρέχεται απευθείας από τον Skyflow ή τον συνεργάτη της πλατφόρμας podcast. Εάν πιστεύετε ότι κάποιος χρησιμοποιεί το έργο σας που προστατεύεται από πνευματικά δικαιώματα χωρίς την άδειά σας, μπορείτε να ακολουθήσετε τη διαδικασία που περιγράφεται εδώ https://el.player.fm/legal.
Player FM - Εφαρμογή podcast
Πηγαίνετε εκτός σύνδεσης με την εφαρμογή Player FM !

Differential Privacy with the University of Victoria’s Dr. Yun Lu

46:31
 
Μοίρασέ το
 

Manage episode 344683816 series 3386287
Το περιεχόμενο παρέχεται από το Skyflow. Όλο το περιεχόμενο podcast, συμπεριλαμβανομένων των επεισοδίων, των γραφικών και των περιγραφών podcast, μεταφορτώνεται και παρέχεται απευθείας από τον Skyflow ή τον συνεργάτη της πλατφόρμας podcast. Εάν πιστεύετε ότι κάποιος χρησιμοποιεί το έργο σας που προστατεύεται από πνευματικά δικαιώματα χωρίς την άδειά σας, μπορείτε να ακολουθήσετε τη διαδικασία που περιγράφεται εδώ https://el.player.fm/legal.

Differential privacy provides a mathematical definition of what privacy is in the context of user data. In lay terms, a data set is said to be differentially private if the existence or lack of existence of a particular piece of data doesn't impact the end result. Differential privacy protects an individual's information essentially as if her information were not used in the analysis at all.

This is a promising area of research and one of the future privacy-enhancing technologies that many people in the privacy community are excited about. However, it's not just theoretical, differential privacy is already being used by large technology companies like Google and Apple as well as in US Census result reporting.

Dr. Yun Lu of the University of Victoria specializes in differential privacy and she joins the show to explain differential privacy, why it's such a promising and compelling framework, and share some of her research on applying differential privacy in voting and election result reporting.

Topics:

  • What’s your educational background and work history?
  • What is differential privacy?
  • What’s the history of differential privacy? Where did this idea come from?
  • How does differential privacy cast doubt on the results of the data?
  • What problems does differential privacy solve that can’t be solved by existing privacy technologies?
  • When adding noise to a dataset, is the noise always random or does it need to be somehow correlated with the original dataset’s distribution?
  • How do you choose an epsilon?
  • What are the common approaches to differential privacy?
  • What are some of the practical applications of differential privacy so far?
  • How is differential privacy used for training a machine learning model?
  • What are some of the challenges with implementing differential privacy?
  • What are the limitations of differential privacy?
  • What area of privacy does your research focus on?
  • Can you talk a bit about the work you did on voting data privacy
  • How have politicians exploited the data available on voters?
  • How can we prevent privacy leakage when releasing election results?
  • What are some of the big challenges in privacy research today that we need to try to solve?
  • What future privacy technologies are you excited about?

Resources:

  continue reading

63 επεισόδια

Artwork
iconΜοίρασέ το
 
Manage episode 344683816 series 3386287
Το περιεχόμενο παρέχεται από το Skyflow. Όλο το περιεχόμενο podcast, συμπεριλαμβανομένων των επεισοδίων, των γραφικών και των περιγραφών podcast, μεταφορτώνεται και παρέχεται απευθείας από τον Skyflow ή τον συνεργάτη της πλατφόρμας podcast. Εάν πιστεύετε ότι κάποιος χρησιμοποιεί το έργο σας που προστατεύεται από πνευματικά δικαιώματα χωρίς την άδειά σας, μπορείτε να ακολουθήσετε τη διαδικασία που περιγράφεται εδώ https://el.player.fm/legal.

Differential privacy provides a mathematical definition of what privacy is in the context of user data. In lay terms, a data set is said to be differentially private if the existence or lack of existence of a particular piece of data doesn't impact the end result. Differential privacy protects an individual's information essentially as if her information were not used in the analysis at all.

This is a promising area of research and one of the future privacy-enhancing technologies that many people in the privacy community are excited about. However, it's not just theoretical, differential privacy is already being used by large technology companies like Google and Apple as well as in US Census result reporting.

Dr. Yun Lu of the University of Victoria specializes in differential privacy and she joins the show to explain differential privacy, why it's such a promising and compelling framework, and share some of her research on applying differential privacy in voting and election result reporting.

Topics:

  • What’s your educational background and work history?
  • What is differential privacy?
  • What’s the history of differential privacy? Where did this idea come from?
  • How does differential privacy cast doubt on the results of the data?
  • What problems does differential privacy solve that can’t be solved by existing privacy technologies?
  • When adding noise to a dataset, is the noise always random or does it need to be somehow correlated with the original dataset’s distribution?
  • How do you choose an epsilon?
  • What are the common approaches to differential privacy?
  • What are some of the practical applications of differential privacy so far?
  • How is differential privacy used for training a machine learning model?
  • What are some of the challenges with implementing differential privacy?
  • What are the limitations of differential privacy?
  • What area of privacy does your research focus on?
  • Can you talk a bit about the work you did on voting data privacy
  • How have politicians exploited the data available on voters?
  • How can we prevent privacy leakage when releasing election results?
  • What are some of the big challenges in privacy research today that we need to try to solve?
  • What future privacy technologies are you excited about?

Resources:

  continue reading

63 επεισόδια

Toate episoadele

×
 
Loading …

Καλώς ήλθατε στο Player FM!

Το FM Player σαρώνει τον ιστό για podcasts υψηλής ποιότητας για να απολαύσετε αυτή τη στιγμή. Είναι η καλύτερη εφαρμογή podcast και λειτουργεί σε Android, iPhone και στον ιστό. Εγγραφή για συγχρονισμό συνδρομών σε όλες τις συσκευές.

 

Οδηγός γρήγορης αναφοράς