1) Loaded validation data to ftp server for participants to access.
2) Fixed docker for evaluation script which was using an old version of sklearn for python. This was necessary to access the Balanced Accuracy Score method of sklearn.
3) Test ran docker submission from Tahsin to simulate a participant. NIH has done the same thing and we got comparable results. NIH should be all set to run these dockers and MGH will be on standby if more resources are necessary.
| Planned for next month |
|---|
1) Setup test data on ftp server 2) Learn docker integrations of caMic and ePAD |
| Comments |
|---|
| We need to tell NIH to reserve two VMs instead of one to run participants submissions. Submissions may take days (2-3). |
| Meeting | Date |
|---|---|
| Bi-weekly Meeting #1 (canceled) |
|
| Bi-weekly Meeting #2 |
|
Milestones
| Description | Date |
|---|---|
| sklearn update in worker docker | 8/15/2019 |
| Identified main evaluation metrics for participants | 8/12/2019 |
Task
| Description | Resolution | Status | Creation Date | Close Date |
|---|---|---|---|---|
| Some confusion on when testing phase submissions are allowed | Working with Tahsin and Spiridon to get clarity | In Progress | 8/31/2019 | -- |
| Traversing notes about caMicroscope/EPad/TCIA integration | In progress | In progress | 8/31/2019 |
Risks
| Description | Mitigation | Rank | Status | Creation Date | Realization | Close Date |
|---|---|---|---|---|---|---|
| Long evaluation time for participant submissions | Multiple VMs on NIH's side as well as MGH resources at the ready. | 1 | Pending | 8/31/2019 | -- | -- |
