DATA MINING techniques. PREDICTIVE MODELS with SAS Enterprise Miner

SAS Institute implements data mining in Enterprise Miner software, which will be used in this book focused predictive models. SAS Institute defines the concept of Data Mining as the process of selecting (Selecting), explore (Exploring), modify (Modifying), modeling (Modeling) and rating (Assessment) large amounts of data with the aim of uncovering unknown patterns which can be used as a comparative advantage with respect to competitors. This process is summarized with the acronym SEMMA which are the initials of the 5 phases which comprise the process of Data Mining according to SAS Institute. The essential content of the book is as follows: SAS ENTERPRISE MINER WORKING ENVIRONMENT MODELLING PREDICTIVE TECHNIQUES WITH SAS ENTERPRISE MINER REGRESSION NODE: MULTIPLE REGRESSION MODEL LOGISTIC REGRESSION DMINE REGRESSION NODE PARTIAL LEAST SQUARES NODE. PLS REGRESSION LARS NODE CLASSIFICATION PREDICTIVE TECHNIQUES. DECISION TREES WITH SAS ENTERPRISE MINER DECISION TREE NODE PREDICTIVE MODELS WITH NEURAL NETWORKS WITH SAS ENTERPRISE MINER OPTIMIZATION AND ADJUSTMENT OF MODELS WITH NETS: NEURAL NETWORK NODE SIMPLE NEURAL NETWORKS PERCEPTRONS HIDDEN LAYERS MULTILAYER PERCEPTRONS (MLPS) RADIAL BASIS FUNCTION (RBF) NETWORKS SCORING AUTONEURAL NODE NETWORK ARCHITECTURES NEURAL NODE TWOSTAGE NODE GRADIENT BOOSTING NODE MEMORY-BASED REASONING (MBR) NODE RULE INDUCTION NODE ENSEMBLE NODE COMBINING MODELS USING THE ENSEMBLE NODE MODEL IMPORT NODE SVM NODE ASSESS PHASE IN DATA MINING PROCESS CUTOFF NODE DECISIONS NODE MODEL COMPARISON NODE SCORE NODE

SAS Institute在Enterprise Miner軟體中實施資料探勘，該軟體將用於本書中的預測模型。SAS Institute將資料探勘定義為選擇（Selecting）、探索（Exploring）、修改（Modifying）、建模（Modeling）和評估（Assessment）大量資料的過程，旨在發現未知模式，並將其用作與競爭對手相比的競爭優勢。根據SAS Institute的說法，這個過程可以總結為SEMMA這個縮寫，該縮寫代表了資料探勘過程中的5個階段。本書的主要內容如下：SAS ENTERPRISE MINER工作環境、使用SAS ENTERPRISE MINER進行建模的預測技術、回歸節點：多元回歸模型、邏輯回歸、DMINE回歸節點、偏最小二乘節點（PLS回歸）、LARS節點、分類預測技術、使用SAS ENTERPRISE MINER的決策樹、決策樹節點、使用神經網絡進行預測模型、使用SAS ENTERPRISE MINER進行優化和調整模型、使用神經網絡節點進行模型評估、簡單神經網絡、感知器、隱藏層、多層感知器（MLPs）、徑向基函數（RBF）網絡、評分、自動神經網絡節點、網絡架構、神經節點、兩階段節點、梯度提升節點、基於記憶的推理（MBR）節點、規則歸納節點、集成節點、使用集成節點組合模型、模型導入節點、支持向量機（SVM）節點、資料探勘過程中的評估階段、截斷節點、決策節點、模型比較節點、評分節點。