Hi There, Do you want us to send results of a leave one out cross validation? Is there any preferences? Thanks, Samad

Created by S J SAJA
Ideally you should avoid overfitting and make your model more generalizable... again the focus of your method should be how to choose the correct subset of genes. P
Dear Pablo, I have the same question! Thanks!
Hi Pablo, As everybody has the GS and we can train and test our model on all data to show better results it is confusing how you will evaluate predictions. I am expecting a standard and fix training and testing of model from you. For example, if organizers saying everybody must submit results of leave one out cross validation makes sense but if you say however you want doesn?t make sense. If really however we want, I will prefer training and testing on all data to show better performance. I am guessing we are missing something here?!
Hi, you can train however you want, as everybody has the GS it is fair. P

Training and testing of predictive model page is loading…