I'm trying to put together a little decision support tool for my department but I am having some trouble with some of the data I am using to build it.
I have some scoring mechanisms for the state of various events and I need to predict the likelihood of a certain outcome. However, outcomes are integer values. So my test data looks like this
0.2 -2
0.14 1
0.82 3
1.3 4
0.62 2
0.12 0
0.22 0
0.54 2
0.04 0
etc.
so in order for this tool to work I need to input a state score (the real numbers) and kick out the likelihood of each outcome state (integers 0 through 5)..
If the outcome states were real numbers (not restricted to integers), then linear regression would make sense. But I don't know enough about statistics to know if linear regression applies in this case.
I have some scoring mechanisms for the state of various events and I need to predict the likelihood of a certain outcome. However, outcomes are integer values. So my test data looks like this
0.2 -2
0.14 1
0.82 3
1.3 4
0.62 2
0.12 0
0.22 0
0.54 2
0.04 0
etc.
so in order for this tool to work I need to input a state score (the real numbers) and kick out the likelihood of each outcome state (integers 0 through 5)..
If the outcome states were real numbers (not restricted to integers), then linear regression would make sense. But I don't know enough about statistics to know if linear regression applies in this case.