james/AIML430_seminar_notes
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Finish seminar 1 notes

Browse Source
This commit is contained in:
1jamesthompson1
2025-08-18 11:11:24 +12:00
parent 574ff09929
commit da7e9dd6ef
12 changed files with 64 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
tue5thaugust/poverty estimates/notes.md → tue5thaugust/tue5thaugust.md
View File

@@ -6,4 +6,9 @@ I think in principle the idea of using earth observation and readily available i
Some questions that I have are:
- I assume that they used some other "off the shelf" style models or datasets to get the earth observation data. However did they consider/investigate the idea of using satalite imagery directly with Neural Networks?
- With regard to the under performance in the urban areas was it due to the data not being dense enough in that area (i.e trying to predict 10,000 people from the same size set of earth observation as opposed to only 100 people in rural areas), or was it more to do with urban areas have differnet relationships with assests and spending so less applicable of a method?
- With regard to the under performance in the urban areas was it due to the data not being dense enough in that area (i.e trying to predict 10,000 people from the same size set of earth observation as opposed to only 100 people in rural areas), or was it more to do with urban areas have differnet relationships with assests and spending so less applicable of a method?
# Carbon emissions
Not completed as it was our group presenting.