|
|
| Line 1: |
Line 1: |
| | + | ==Mary Gates== |
| | February 24 - Application Due for Mary Gates Research Presentation | | February 24 - Application Due for Mary Gates Research Presentation |
| | | | |
| | May 19 - Research Symposium | | May 19 - Research Symposium |
| | | | |
| − | So I also need to put in milestones for things besides these required MG responsibilities.
| + | ==General== |
| − | Basiclly I should be collecting data constantly and need to be building up my data set of users posting on certain topics which requires slightly more automation of the current process that I have of collected data/analyzing for bursts.
| + | * Models for transmission graph (Learning, Prediction) |
| − | | + | * Precition <-> Recall Graphs on predictions |
| − | The second step will be using those transmissions to build a probabilistic model for the transmission graph. This also appears to have been previously done in 'Information Diffusion'. The model will then be learned up the source of data.
| + | * Compare Models |
| − | | |
| − | Then I will need to conduct some more fulfilling tests than has been done previously, hopefully either in predicting the magnitude of the eventual spread or in showing that my predictions for the likelihood of which user will transmit the topic next are accurate. How is this to be done?
| |
| − | | |
| − | Given a set of tests it would then even be possible to compare the different models for transmission rather than just confirming that the model determines the most influential users.
| |
| − | | |
| − | More to come after I listen to some talks!
| |
Revision as of 00:47, 2 February 2006
Mary Gates
February 24 - Application Due for Mary Gates Research Presentation
May 19 - Research Symposium
General
- Models for transmission graph (Learning, Prediction)
- Precition <-> Recall Graphs on predictions
- Compare Models
