I think it wasn’t that computers were to slow, but rather getting yelled at by statisticians if the number of training parameters was bigger than the sample (training dataset) size
Training Parameters vs Sample Size: Statisticians’ Historical Concerns
By
–
Leave a Reply