George Fernandez, currently employed as a statistical training specialist at SAS Institute. Before joining SAS, he worked as a Professor of applied statistics, and the Director for the University of Nevada- Reno Center for Research Design and Analysis and has more than 23 years of experience in teaching courses such as introductory statistical methods, design and analysis of experiments, linear and non-linear regression, multivariate statistical methods and SAS programming. He is a professional SAS programmer and has over 30 years’ experience in many analytical and graphical SAS modules. He has won best paper and poster presentation awards at the regional and international conferences. He has presented several invited full-day workshops on "Applications of user-friendly statistical methods in Data mining: American Statistical Association Joint meeting in Atlanta (2001), Western SAS users Conference in Arizona (2000), in San Diego (2002),and San Jose (2005), 56th Deming's conference, Atlantic City (2003), Key-note Speaker and workshop presenter, 16th Conference on Applied Statistics, Kansas State University. He has also organized 7th Western Users of SAS conference (WUSS) at Los Angeles in 1999 and served as the section chair, SUGI31 and SGF2007-2009. His book on "Statistical Data mining using SAS applications" (2ed CRC press / Chapman Hall) contains many user-friendly SAS macro-applications.
One of the primary predictive modeling tools capable of fitting very complex nonlinear functions is Neural Networks (NN) in SAS advanced analytics software SAS Enterprise Miner (EM). The default option in NN EM node uses a Multilayer Perceptron model with an input, an output and a single hidden layer with
