National Repository of Grey Literature 1 records found  Search took 0.00 seconds. 
Long steps in IPM and L_1-regression
Šicková, Barbora ; Černý, Michal (advisor) ; Sokol, Ondřej (referee)
This work deals with Newton method of interior point which is applied on finding estimation of polynomial L_1 regression. The aim of this thesis is to find new modifications of long step choice in Newton method in order to find faster solution of L_1 estimations for large data sets. Designed modifications are based on full-Newton step algorithm for the self-dual model. According to reasults, the best are algorithms AF-L, F-LP1, AF-LP1 a AF-L-mixed, which modify the barrier update parameter in adaptive way. Algorithms and obtained results were implemented end visualized in MatLab.

Interested in being notified about new results for this query?
Subscribe to the RSS feed.