I don't have the statistical background to describe this, but here goes....
I am developing a system to track historical data in order to have it
develop a dynamic model to be used for forecasting. For example, about 5
input variables can pretty well predict the outcome variable (I'm probably
not using the right terms here). The relationship will never be exact, and
I will need to determine which input variables have the most impact on the
sensitivity of the outcome. I will have access to all historical input and
outcome data. I will also have access to forecasts for the input variables,
so the idea is to develop a prediction for the outcome using the model.
Can anyone suggest a good place to start on this problem. Are there any
perl libraries or examples that would help?
Jarson
I am developing a system to track historical data in order to have it
develop a dynamic model to be used for forecasting. For example, about 5
input variables can pretty well predict the outcome variable (I'm probably
not using the right terms here). The relationship will never be exact, and
I will need to determine which input variables have the most impact on the
sensitivity of the outcome. I will have access to all historical input and
outcome data. I will also have access to forecasts for the input variables,
so the idea is to develop a prediction for the outcome using the model.
Can anyone suggest a good place to start on this problem. Are there any
perl libraries or examples that would help?
Jarson