I have a numeric dataset in the following format
X1 X2 output(Y)
1 2 3
2 5 1
2 3 2
2 2 2 . . .I am looking for a way to model the data and find the weights (b0,b1,b2)
Y = b0 + b1X1 + b2X2One way to do this is to use, linear Regression for Multiple Variables.
http://www.nd.com/NSBook/NEURAL%20AN...ple_Varia.html
http://www.statsoft.com/Textbook/Multiple-Regression
But I wonder if there is a better way to build a model? For example, we can build a model and know the weights such as Y = 1.3433 + 0.2836*x1 + 0.004*x2But the values 1.3 or 0.28 don't say much about the relations in the data. There should be better ways to find weights and build models. Something that have summations or divisions, square root, etc, such as
Y = b0 + ∑b1*x1 + b2*x2^2
Any suggestions that can Weka help with ?
Thanks!
X1 X2 output(Y)
1 2 3
2 5 1
2 3 2
2 2 2 . . .I am looking for a way to model the data and find the weights (b0,b1,b2)
Y = b0 + b1X1 + b2X2One way to do this is to use, linear Regression for Multiple Variables.
http://www.nd.com/NSBook/NEURAL%20AN...ple_Varia.html
http://www.statsoft.com/Textbook/Multiple-Regression
But I wonder if there is a better way to build a model? For example, we can build a model and know the weights such as Y = 1.3433 + 0.2836*x1 + 0.004*x2But the values 1.3 or 0.28 don't say much about the relations in the data. There should be better ways to find weights and build models. Something that have summations or divisions, square root, etc, such as
Y = b0 + ∑b1*x1 + b2*x2^2
Any suggestions that can Weka help with ?
Thanks!