Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Recording Prediction Error for Travel and Dwell Times Seperately #225

Open
simonberrebi opened this issue Feb 13, 2021 · 2 comments · May be fixed by #226
Open

Recording Prediction Error for Travel and Dwell Times Seperately #225

simonberrebi opened this issue Feb 13, 2021 · 2 comments · May be fixed by #226
Assignees
@scrudden
Labels
enhancement

Comments

@simonberrebi
Copy link
Member

Measuring prediction error separately would allow to benchmark how travel and dwell times contribute to prediction errors. This benchmarking could inform the prioritization of new development efforts. The benchmarking could also assess the performance of prediction methods and help identify bugs. Recording prediction errors for constituent parts separately is therefore necessary to guide the development of prediction methods addressing dwell time and travel time specifically.
@scrudden scrudden self-assigned this Feb 13, 2021
@scrudden scrudden linked a pull request Feb 13, 2021 that will close this issue
Record values informing error for dwell and travel times separately. #226
Open
@scrudden
Copy link
Member

Code on branch tc_issue_225

@simonberrebi
Copy link
Member Author

Based on the new data fields developed by @scrudden, we are now able to record prediction errors for travel and dwell times separately. The following graph shows 12.5 to 87.7 percentiles of dwell and travel prediction errors against the horizon on the MBTA's Route 66 in Boston. We can see that dwell time predictions tend to be overly optimistic whereas travel time predictions are pessimistic. They cancel each other out. This shows the value of considering travel and dwell time predictions separately.

Prediction Accuracy by Component

@scrudden scrudden mentioned this issue Mar 24, 2021
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@scrudden scrudden

Labels
enhancement
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Record values informing error for dwell and travel times separately. TheTransitClock/transitime
2 participants
@scrudden @simonberrebi