本研究以專利分析法探討美國半導體產業公司市值,並透過類神經網絡的運算法則分析原告次數專利熵指數(Entropy)、研發支出費用(R&D)對公司市場價值的影響,專利分析與財務的資料來源分別為美國專利商標局[USPTO]和[COMPUSTAT],法務相關資訊則由[WESTLAW INTERNATIONAL]所提供。彙整成資料庫做原始資料。透過倒傳遞類神經網絡的方式做分析,其研究結果發現原告次數與公司市值表現呈現出正向的關係。專利熵指(Entropy)在一定的集中度後會有大幅度的正向關係成長。因此,本研究有助於釐清美國半導體產業專利投資方向與實證集中一相關領域技術的必要性假說,藉此研擬策略並強化競爭優勢。
This study uses patent analysis to explore the influences of R&D expenditures, patent litigation, and entropy upon net income in the American semiconductor industry by artificial neural networks. By using the multilayer perceptron, the results show that Patent Litigation positively affect net income in the American semiconductor companies, while entropy positively affects it in these companies. Therefore, This study will help the US semiconductor industry patent investment direction and positive focus hypothesis of a related art whereby to maintain a competitive advantage, the performance of robust market capitalization.