Thanks for your great topic and interesting model. However, I have a question of the SAN model. When proving the generalizability of the SAN model, you use the data of children and adolescents to prove that the SAN model can predict symptoms of ADHD. Nevertheless, for people who are suffering from ADHD, a huge part of them are people who are from a stroke or from chronic traumatic encephalopathy. Have you tried to use data from these people to prove the efficiency of the SAN model? Will the change of data influence the model's performance?
Hi Monica,
Thanks for your great topic and interesting model. However, I have a question of the SAN model. When proving the generalizability of the SAN model, you use the data of children and adolescents to prove that the SAN model can predict symptoms of ADHD. Nevertheless, for people who are suffering from ADHD, a huge part of them are people who are from a stroke or from chronic traumatic encephalopathy. Have you tried to use data from these people to prove the efficiency of the SAN model? Will the change of data influence the model's performance?
Thanks again for this inspiring topic.