[Next] [Previous] [Top] [Contents]
Towards a Descriptive Model of Agent Strategy Search - Bruce Edmonds
7. Discussion
Clearly the model exhibited represents only a step towards an accurate description of the target behaviour observed by Sonnemans. The guidance from cognitive science as to the nature of the learning processes involved is, at present, only suggestive. This results in there being considerable uncertainty about the nature of the learning process, uncertainty that the current observational data does not completely resolve. What the model does do is to point up what information we are lacking about the target behaviour that would enable us to improve the model. In this case one thing that is missing is any direct information about how the subjects are developing their models as they learn. This could be approached in several ways, including:
- Getting the subjects to specify their strategies using Sonneman's techniques at several stages throughout the learning process, (including before they start);
- The subjects could be asked to specify several strategies under the condition that they would receive the earnings gained by the best performing strategy over a series of runs - this would give a picture of the variety of strategies they were considering;
- Constraining the development of strategies by the subjects to combinations of the operators in the above model using the same `electronic information board' techniques that Sonnemans uses to capture the strategies, so that some idea about the sort of manipulation and parameterisation of the strategies could be gained;
- Keeping and making available the choice that the subjects made over time so that the dynamics of the learning processes could be checked.
If these were done the process of descriptive modelling could go through another, more accurate iteration. This might highlight more facets that would require experimental or observational investigation. Alternatively it might reveal that some of the inevitable assumptions we had made about the learning process were unwarranted. But in either case we would have learnt something about the actual learning processes involved.
Towards a Descriptive Model of Agent Strategy Search - Bruce Edmonds - 06 SEP 99
[Next] [Previous] [Top] [Contents]
Generated with CERN WebMaker