2

Quasi-newton method： If the second derivative of the target function doesn't exist at all, can we use Quasi-newton method? If the second derivative of the target function at some point doesn't exist, can we use Quasi-newton method? My question comes from the fact that the second derivative of the SCAD penalty function doesn't exist at some point——http://www.personal.psu.edu/ril4/research/penlike.pdf

Please forgive me, my English is not very good, thank you

David G. Stork
  • 29,774
abraxas
  • 21
  • Welcome to stackexchange. The convention here is one question per post. In fact, your first question probably belongs on https://stats.stackexchange.com/. Your second question points to a long technical paper. Please [edit] the question to show us just the part that you are asking about. PS your English is fine. – Ethan Bolker Feb 16 '19 at 03:33
  • @EthanBolker Our Dear Ethan~Thank you。Never mind that paper, the main question I want to ask is whether we can use quasi-newtonian methods when the second derivative doesn't exist。Also, I am a novice, how can I find the answer to the first question in the link you gave? – abraxas Feb 16 '19 at 03:36
  • 1
    The link I gave is to the statistics stackexchange. You can't find the answer to your second question there - you should ask it there instead of here since readers there are more likely to be able to help you. Leave the second question in this post when you [edit]. – Ethan Bolker Feb 16 '19 at 03:44
  • @EthanBolker Our Dear Ethan，I edited it as you asked – abraxas Feb 16 '19 at 03:49

0 Answers0