RE: [maths] Lagrange Multipliers & Optimal operation of ....

["Anh Le" <anh_t_le@hotmail.com>]

Cha`o ca'c ba'c,

  Ba'c TD no'i la`m to^i to` mo` qua'. Cha('c lu'c ra?nh pha?i mu+o+.n 
quye^?n sa'ch ddo' ve^` ddo.c xem co' hie^?u ddu+o+.c kho^ng.
  To^i ca?m tu+o+?ng nhu+ Dimitri Bertsekas chuye^n tri. vie^'t sa'ch gia'o 
khoa kho' hie^?u va` kho^ng thu+.c te^'. Ma^'y o^ng ddo' nha^.n Grants tha? 
gia`n cu?a US Airforce, US Navy..., nghie^n cu+'u theo kie^?u cuo^'n chie^'u 
ro^`i vie^'t tha`nh sa'ch gia'o khoa luo^n. Tha`nh ra sa'ch vie^'t kho^ng 
sai va`o dda^u ddu+o+.c, co' the^? du`ng la`m tra cu+'u to^'t, nhu+ng vo+'i 
beginners thi` kho' hie^?u, ddo^'i vo+'i seniors thi` kho^ng co' gia' tri. 
thu+.c te^'.
  To^i co' tu+`ng ddo.c cuo^'n Data networking, va` cuo^'n Linear network 
optimization cha(?ng ha.n, kho^ng ngo+` nhu+~ng thu+' ddo+n gia?n mi`nh ho.c 
ho^`i xu+a nhu+ simplex method, ford-folkerson, Little theorem, Jackson 
theorem...la.i bie^'n tha`nh nhu+~ng co^ng thu+'c nghie^m chi?nh va` ra('c 
ro^'i the^' (hay do la^u ro^`i to^i kho^ng ddo.c gio+` tha^'y co^ng thu+'c 
la` cho'ng ca? ma(.t-:()
   Trong khi ddo' va`i mo^ hi`nh ddo+n gia?n ho+n ne^n hay u+'ng du.ng 
ddu+o+.c nhu+ Kelly network (thay cho Jackson) la.i kho^ng no'i qua, integer 
linear programming hi`nh nhu+ toa`n tha^'y thie^n ha. gia?i ba(`ng ca'i 
kie^?u heuristics tu+. nghi~ (*), simplex method thi` to^i va^~n thi'ch 
ca'ch gia?i thi'ch hi`nh ho.c trong mo^.t quye^?n tie^'ng Nga cu~ ky~ ho+n 
(**).
   So, chuye^'n na`y pha?i ti`m tho+`i gian xem quye^?n mo+'i na`y. Ne^'u 
va^~n tha^'y tha^'y co`n mo^ng lung nhu+ ma^'y quye^?n kia, e ra(`ng pha?i 
bo+'t ta'n do'c ma` o^n la.i kie^'n thu+'c cu~ ma^'t-:).

Cheers,
Tuan Anh.
(*) Nghe no'i ca'c microprocessors hie^.n dda.i cu~ng vi` ma^'y chuye^.n 
TSP, integer linear programming na`y ma` le.t dde.t. Bo.n ho. toa`n ta(ng 
clock frequency dde^? impress kha'ch ha`ng chu+' hie^.u qua? cha(?ng 
ddu+o+.c bao nhie^u. Ta(ng MHz dde^? ro^`i la`m nhie^`u le^.nh trong cu`ng 
mo^.t nhi.p, co' algorithm ddoa'n tru+o+'c ~ le^.nh na`o se~ thu+.c hie^.n 
ke^' tie^'p va` ddo^.c la^.p dde^? load va`o ngay. Xe't cho cu`ng cu~ng la` 
mo^.t kie^?u ti`m kie^'m theo "kinh nghie^.m", ddoa'n ddu'ng thi` ddo+~, 
lo+~ ddoa'n sai thi` la~nh ddu? vi` pha?i la`m la.i tu+` dda^`u-:)
(**) Quye^?n ddo' to^i nho+' la` cu?a Zukhovisky, cha('c nhie^`u ba'c bie^'t 
(co' refers dde^'n mo^.t ba`i ba'o cu?a GS Hoa`ng Tu.y na(m 1964). Cu~ng co' 
the^? lu'c ddo' dda^`u o'c co`n ra?nh rang ne^n ddo.c ca'i gi` cu~ng va`o 
a`o a.t-:)

From: "Hoang Duong Tuan" <tuan@toyota-ti.ac.jp>

>Theo to^i quye^?n na`y vie^'t tu+o+ng ddo^'i modern, de^~ hie^?u va` mu.c
>ddi'ch dde^? solve optimization problems thu+.c su+. chu+' kho^ng pha?i
>dde^? tre^n tro+`i du+o+'i bie^?n nhu+ nhie^`u quye^?n nonlinear progamming
>kha'c.
>cheers,
>TD


_______________________________________________________________
Get Free Email and Do More On The Web. Visit http://www.msn.com