> Actually, in my work, I'll use Newton's method which involves many > matrix multiplication and subtraction. It seems that you have finished > this part.
Yes, It done. > BTW, although I plan to use Quasi-Newton method to avoid calculate the > inverse of Hessian directly, I'm wonder do you plan to implement the > inverse operation? Of course, We also have a lot of needs and issues regarding inverse operation. Regards, Edward J. Yoon On Wed, Jul 23, 2008 at 11:02 PM, Yun Jiang <[EMAIL PROTECTED]> wrote: > Hi, > That's great, since this can reduce much of my work:) > Actually, in my work, I'll use Newton's method which involves many > matrix multiplication and subtraction. It seems that you have finished > this part. > > BTW, although I plan to use Quasi-Newton method to avoid calculate the > inverse of Hessian directly, I'm wonder do you plan to implement the > inverse operation? > > Best, > > On Wed, Jul 23, 2008 at 9:28 PM, Chanwit Kaewkasi <[EMAIL PROTECTED]> wrote: >> Hello Yun, >> >> Glad to see another gsoc-er around :-) >> I'm doing gsoc for another project (Groovy) while helping Edward >> incubate this one. >> >> Cheers, >> >> Chanwit >> >> 2008/7/23 Yun Jiang <[EMAIL PROTECTED]>: >>> Hi, all >>> I'm implementing Logistic Regression in project of mahout in gsoc. >>> Is that ok if I use hama in my project? >>> >>> Thanks, >>> Yun >>> >> > -- Best regards, Edward J. Yoon [EMAIL PROTECTED] http://blog.udanax.org
