- 學位論文
轉換型模糊C-均值演算法應用於皮膚病與鳶尾花的分群
Applying Fuzzy Transformed C-Means Algorithm to the Dermatology Disease and Iris Clustering
摘要
在台灣溫暖潮濕的氣候型態下，皮膚病是人們常會發生的疾病，而發病的原因相當多，不同的皮膚病會有不同的症狀。本研究以六種皮膚疾病資料進行實驗，也以鳶尾花資料同步進行實驗，比較下列三種目標函數導向分群方法之分群準確度。 Bezdek於1981年提出眾所皆知的模糊C-均值算法（ FCM ）是一種基於目標函數的分群方法。因此，不同的目標函數可能會導致不同的結果。重要的問題是如何獲得一個更緊湊與可分離的目標函數，以改善集群準確性。吳國隆等於2005年提出的FCS目標函數是一個以結合模糊內部和群間的差異來改進FCM目標函數的優良演算法。劉湘川於2009年提出改良的轉換型模糊C-均值演算法（ FTCM ）可獲得更多的可分離數據轉換。實驗結果顯示FTCM演算法的表現優於傳統的FCM演算法和FCS演算法。 藉由此研究，若將這些人工智慧的分群技術，提供醫師作為診斷時的輔助參考，或患者就醫前的自我檢核，將可避免不必要的醫療資源浪費，以提升醫療服務品質。
並列摘要
Dermatological disease is common in places that are humid,damp, and hot like Taiwan, various reasons can cause skin diseases with which the symptoms are varied. In this study, six types of skin diseases experimental data, iris data synchronization also conduct experiments to verify the accuracy of the experiment. The popular fuzzy c-means algorithm (FCM) proposed by Bezdek in 1981is an objective function based clustering method. Hence, different objective function may lead to different results. The important issue is how to get a more compact and separable objective function to improve the cluster accuracy. The objective function of the well known improved algorithm, FCS proposed by G. L. Lee et al in 2005, is a generalization of the FCM objective function by combining fuzzy within- and between-cluster variations. In this paper, a more separable data transformation, the improved new algorithm, “Fuzzy Transformed C-Means algorithm(FTCM)”, is proposed by Hsiang-Chuan Liu in 2009. Two real data sets were applied to prove that the performance of the FTCM algorithm is better than the conventional FCM algorithm and the FCS algorithm. In case study, artificial intelligence technology was able to provide a diagnostic tool for physician or to help patients with self-check before treatment; it could avoids unnecessary waste of medical resources and improves the quality of medical services.