Birch algorithm . density-based method : if the density of neighborhood, that is the number of data objects, exceeds a certain value, the clustering process will be continued 基于密度的方法的主要思想是:只要鄰近區(qū)域的密度(對象或數(shù)據(jù)點(diǎn)的數(shù)目)超過某個(gè)閾值,就繼續(xù)聚類。
Dissimilarities are assessed base on the attribute values describing the objects . clustering processes are always carried out in the condition without pre-known knowledge, so the main task is to solve that how to get the clustering result in this premise 聚類分析依據(jù)的原則是使同一聚簇中的對象具有盡可能大的相似性,而不同聚簇中的對象具有盡可能大的相異性,聚類分析主要解決的問題是如何在沒有先驗(yàn)知識的前提下,實(shí)現(xiàn)滿足這種要求的聚簇的聚合。
However, there have still some unresolved problems : first, how to determine the number and size of the clusters automatically during the clustering process . second, how to utilize the " local " ridge regression method which including multiple regularization parameters in learning rbf network . third, those clusters in irregular form ca n't represented by radial basis function, thus we must find some other basis functions that can describe the irregular form 但是仍然存在幾個(gè)問題尚待解決:首先,聚類時(shí)怎樣自動確定簇的個(gè)數(shù)和半徑;其次,如何利用含有多個(gè)正規(guī)化參數(shù)的局部嶺回歸方法進(jìn)行rbf網(wǎng)絡(luò)學(xué)習(xí);第三,如果簇的形狀是不規(guī)則的,則它很難用徑向基函數(shù)來描述,因此需要研究其它能代表不規(guī)則形狀的簇的基函數(shù)。
Clustering analysis is the method which partition class to the clustered objects as required of thing's characteristics . clustering processes are always carried out in the condition with no pre-known knowledge, so the most research task is to solve that how to get the clustering result in this premises 聚類分析是在沒有先驗(yàn)知識支持的前提下,根據(jù)事物本身的特性研究被聚類對象的類別劃分,實(shí)現(xiàn)滿足這種要求的類的聚合,它所依據(jù)的原則是使同一類中的對象具有盡可能大的相似性,而不同類中的對象具有盡可能大的差異性。
clustering processes are always carried out in the condition with no pre-known knowledge, so the most research task is to solve that how to get the clustering result in th is premise . as the development of data mining, a number of clustering algorithms has been founded, in general, major clustering methods, can be classified into the following categories : partitioning methods; hierarchical methods; density-based methods; grid-based methods; model-methods; besides these, some clustering algorithms integrate the ideas of several clustering methods 正是由于聚類分析的重要性和特殊性,近年來在該領(lǐng)域的研究取得了長足的發(fā)展,涌現(xiàn)出了許多聚類分析的方法,如劃分聚類方法(partitioningmethod)、層次聚類方法(hierarchicalmethod)、基于密度(density?based)的聚類方法、基于網(wǎng)格(grid?based)的聚類方法、基于模型(model?based)的聚類方法等等。
And it adds a-priori information into the patterns to change the method as a semi-supervised clustering . in the clustering process, the unlabelled patterns compare similarities with the labeled patterns, and then the accuracy of the algorithm can be increased . ( 3 ) the paper proposes an interactive learning-based image mining in remote sensing 由于遙感圖像各類別在特征空間中散點(diǎn)圖的分布的特點(diǎn),本文對傳統(tǒng)的fcm聚類算法進(jìn)行改進(jìn),并且加入先驗(yàn)信息之后,將原來的非監(jiān)督的聚類變成一種半監(jiān)督的聚類方法,通過與已標(biāo)簽的樣本進(jìn)行相似性比較,能有效地提高聚類算法的準(zhǔn)確度。