How good are we at automatically finding these mistakes?

ProjectSidewalk / sidewalk-quality-analysis

An analysis of Project Sidewalk user quality based on interaction logs

5 stars 3 forks source link

How good are we at automatically finding these mistakes? #21

Open jonfroehlich opened 5 years ago

jonfroehlich commented 5 years ago

Based on our team meeting today, we thought it would be useful for @nchowder to see this chart of common false positives that we found in our Project Sidewalk DC dataset (as described in our CHI'19 paper).

nch0w commented 5 years ago

Do we want to train models to pick these out specifically? I don't know if this is possible using only position data.