A Preliminary Study on Patent Pricing Model Constructed by Nonlinear Interval Regression

• Main Authors: Sung-Lin Tsai, 蔡松霖
• Other Authors: CHIU YU-JING
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/95152232402230500817

碩士 === 中原大學 === 企業管理研究所 === 96 === Abstract Patent right can make a lot of profit for its owner. The source of profit is from patent licensing and damage ward of infringement litigations for compensate. The purpose of the study is to examine the relationship among the corporations owned the value of patent and the reparation patent of litigation adjudge. Moreover, the study uses statistics to predict patent litigation adjudge reparation’s outcome, and provided discussion, practical suggestions and implications based on the results. This study applies neural network internal regression method. This can improve the use of restrictions from fuzzy regression in non-linear problem. Most of the final appraisal indicators are higher than average number and can yield higher amount of verdict, which is of higher patent value. On the other hand, if the appraisal indicators are lower than average number, it can yield lower amount of verdict, which is of lower patent value. If Taiwanese company can apply patent right adequately, they can gain considerable revenue. In addition, enterprises can get high amount compensation of damages through patent litigation either by selling new products or by licensing to get royalty. Keywords: Patent value, neural network internal regression, fuzzy regression.