數(shù)據(jù)挖掘概念與技術(shù)數(shù)據(jù)預(yù)處理-全文預(yù)覽
【正文】 2020/9/15 13 數(shù)據(jù)歸約 歸約數(shù)據(jù)集小、近似保持原數(shù)據(jù)的完整性 數(shù)據(jù)壓縮：編碼機(jī)制壓縮數(shù)據(jù) 方法：小波變換、主成分分析（ PCA） 數(shù)值壓縮：用替代的、較小的數(shù)據(jù)表示數(shù)據(jù) 方法：有參方法 回歸和對(duì)數(shù)線性模型 無(wú)參方法 直方圖 聚類 抽樣 2020/9/15 14 離散化和概念分層 原始值用區(qū)間值或較高 層的概念替換 離散化：原始值劃分為有限個(gè)區(qū)間，減少了連續(xù)屬性值的個(gè)數(shù) 概念分層：原始值用較高層的概念替換 2020/9/15 15 離散化和概念分層： 對(duì)數(shù)值數(shù)據(jù) ：分箱 直方圖分析 聚類分析 基于熵的離散化 (entropy) 通過(guò)信息增益方法確定區(qū)間邊界 使用類信息，更可能將區(qū)間邊界定義準(zhǔn)確 基于卡方分析的區(qū)間合并 通過(guò)自然劃分分段 345規(guī)則 (P58)：將數(shù)值數(shù)據(jù)劃分成相對(duì)一致的區(qū)間 2020/9/15 16 對(duì)分類數(shù)據(jù) : ? 分類數(shù)據(jù)是離散數(shù)據(jù)。customer表的 ine,age屬性。挖掘相關(guān)數(shù)據(jù)可以寫(xiě)成 ： use database AllElectronics_db in relevance to , , , from customer C , item I , purchase P , item_sold S where = and = and = group by 挖掘知識(shí)類型的語(yǔ)法 ? Mine_Knowledge_Specificaton::=Mine_Char | Mine_Discri | Mine_Assoc | Mine_Class | Mine_Pred ? Mine_Char::= Mine characterization [as pattern_name] analyze meansure(s) ? 例： mine characteristics as customerPurchasing analyze count% ? Mine_Discri::=Mine parison [as pattern_name] for target_class where target_condition { versus contrast_class_i where contrast_condition_i} analyze measure(s) ? 例： mine parison as purchaseGroups for bigSpenders where avg() ≥$100 analyze count versus budgetSpenders where avg(),$100 挖掘知識(shí)類型的語(yǔ)法 (續(xù) ) ? Mine_assoc::=mine association [aspattern_name] [matching metapattern] ? 例： mine associations as buyingHabits matching P(X:customer,W)^Q(X,Y)=buys(X,Z) ? Mine_class::=mine classification [as pattern_name] analyze classifying_attribute_or_dimention ? 例： mine classification as classifyingCustomerCreditRating analyze credit_info ? Mine_predi::=Mine prediction [as pattern_name] analyze prediction_attribute_or_dimention {set { attribute_or_dimention_i=value_i}} ? 例： mine prediction as predictItemPrice a
環(huán)評(píng)公示相關(guān)推薦
鄂ICP備17016276號(hào)-1