- From: Justin Li via GitHub <sysbot+gh@w3.org>
- Date: Tue, 12 Jul 2022 13:21:45 +0000
- To: public-patcg@w3.org
Thanks @benjaminsavage and @eriktaubeneck for the clarifications! I was starting to arrive to the same conclusion after inspecting the current state of this repo: [raw-ipa](https://github.com/martinthomson/raw-ipa). I do have 2 follow up questions if you would indulge me. 1. With the global limit on what a trigger values can be, how do we deal with trigger values that will be out of bound? Would it make sense for the browser that is generating the trigger event with an out of bound trigger value to split the event in multiple individual trigger events, each with a trigger value that satisfies the global bound, but still sums up to the original value (basically additive secret share)? 2. If the trigger values are very small that represent for example the number of clicks/app installs, etc. Wouldn't it be possible for the differential private noise computed based on the global bound (much larger than the trigger values) to skew the final aggregate value and lower the over accuracy of the result? -- GitHub Notification of comment by juanli16 Please view or discuss this issue at https://github.com/patcg/private-measurement/issues/9#issuecomment-1181754272 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Tuesday, 12 July 2022 13:21:47 UTC