I am looking for an information on MVR using Microsoft Excel and also what are the differences between using Multi-Variables Regression (MVR) and Artificial Neural Network (ANN) [if there any comparison between MVR and ANN]
It depends a lot on the details, but in general I found ANN to be very intuitive and comfortable for fitting nonlinear functions. If a linear model is sufficient I would probably recommend standard multiple regression as it may be easier to understand (but of course you can use ANN to do linear MVR as well -it all depends on your prior knowledge and on what you want to find in your data). If you need Excel code for MVR or ANN just send me a message
Thanks Andreas
I am using ANN as a prediction method for solid-liquid permeability [and I believe that it is a non-linear data] in order to make value of using the ANN I need to compare ANN results with another statistical program such as MVR in order to answer the question that “ ANN and Excel, which is better to use for prediction? And why? “. If you have or you know any work related to that work, would you please send me. Also, I am happy to receive your both codes for ANN and MVR to check my work and results.
Kind regards
Sorry, the question that “ ANN and MVR, which is better to use for prediction? And why?
MVR regression will try to fit a particular function, such as, linear. ANNs can fit nonlinear functions, although the form the function may not be clear. To test for a particular application, you could take some sample test data and run them both to get an error estimate. Of'course on the internet there are several resources, as mentioned by others that can help you in doing MVR in excel as well as ANNs. Now, there are many techniques under each method. If you hve Matlab or maybe R that may help you in running some tests. Hoep this helps a little. Good Luck.
- Badges
- Science topic