Based on the classical fuzzy cmeans ( fcm ), a function for measuring clustering validity and a heuristic method to calibrate the fuzzy exponet iteratively are defined, and a practical fuzzy neural network model is proposed 在傳統(tǒng)的模糊c-均值算法的基礎(chǔ)上,給出了一個(gè)衡量聚類有效性的函數(shù)和確定模糊指數(shù)的啟發(fā)式方法,并給出了應(yīng)用該算法的具體的模糊神經(jīng)網(wǎng)絡(luò)模型。
Based on fuzzy clustering algorithm, we studied the objective function of the traditional fuzzy c-means algorithm and proposed a modified objective function for fcm; we discussed clustering validity problem, and a texture segmentation method based on adaptive fcm has been constructed by the guidance of fuzzy clustering validity 在介紹聚類算法的基礎(chǔ)上,研究了模糊c-均值聚類算法目標(biāo)函數(shù)的改進(jìn)問題,提出了基于修正目標(biāo)函數(shù)的fcm算法;討論了聚類有效性問題,在模糊聚類有效性函數(shù)指導(dǎo)下構(gòu)造了一種自適應(yīng)模糊c-均值聚類算法的紋理分割方法。
Based on fuzzy clustering algorithm, we studied the objective function of the traditional fuzzy c-means algorithm and proposed a modified objective function for fcm; we discussed clustering validity problem, and a texture segmentation method based on adaptive fcm has been constructed by the guidance of fuzzy clustering validity 在介紹聚類算法的基礎(chǔ)上,研究了模糊c-均值聚類算法目標(biāo)函數(shù)的改進(jìn)問題,提出了基于修正目標(biāo)函數(shù)的fcm算法;討論了聚類有效性問題,在模糊聚類有效性函數(shù)指導(dǎo)下構(gòu)造了一種自適應(yīng)模糊c-均值聚類算法的紋理分割方法。