|
" Effective CRM using predictive analytics / "
Antonios Chorianopoulos
| Document Type
|
:
|
BL
|
| Record Number
|
:
|
662495
|
| Doc. No
|
:
|
dltt
|
| Main Entry
|
:
|
Chorianopoulos, Antonios.
|
| Title & Author
|
:
|
Effective CRM using predictive analytics /\ Antonios Chorianopoulos
|
| Page. NO
|
:
|
1 online resource
|
| ISBN
|
:
|
9781119011569 (Adobe PDF)
|
|
|
:
|
: 1119011566 (Adobe PDF)
|
|
|
:
|
: 9781119011576 (ePub)
|
|
|
:
|
: 1119011574 (ePub)
|
|
|
:
|
9781119011552 (cloth)
|
|
|
:
|
: 9781119011583
|
|
|
:
|
: 1119011582
|
|
|
:
|
: 1119011558 (cloth)
|
|
|
:
|
: 9781119011552 (cloth)
|
| Bibliographies/Indexes
|
:
|
Includes bibliographical references and index
|
| Contents
|
:
|
An overview of data mining: The applications, the methodology, the algorithms, and the data -- The applications -- The methodology -- The algorithms -- Supervised models -- Classification models -- Estimation (regression) models -- Feature selection (field screening) -- Unsupervised models -- Cluster models -- Association (affinity) and sequence models -- Dimensionality reduction models -- Record screening models -- The data -- The mining datamart -- The required data per industry -- The customer "signature": from the mining datamart to the enriched, marketing reference table -- Summary -- The Methodology -- Classification modeling methodology -- An overview of the methodology for classification modeling -- Business understanding and design of the process -- Definition of the business objective -- Definition of the mining approach and of the data model -- Design of the modeling process -- Defining the modeling population -- Determining the modeling (analysis) level -- Definition of the target event and population -- Deciding on time frames -- Data understanding, preparation, and enrichment -- Investigation of data sources -- Selecting the data sources to be used -- Data integration and aggregation -- Data exploration, validation, and cleaning -- Data transformations and enrichment -- Applying a validation technique -- Split or Holdout validation -- Cross or n-fold validation -- Bootstrap validation -- Dealing with imbalanced and rare outcomes -- Balancing -- Applying class weights -- Classification modeling -- Trying different models and parameter settings -- Combining models -- Bagging -- Boosting -- Random Forests -- Model evaluation -- Thorough evaluation of the model accuracy -- Accuracy measures and confusion matrices -- Gains, Response, and Lift charts -- ROC curve -- Profit/ROI charts -- Evaluating a deployed model with test-control groups -- Model deployment -- Scoring customers to roll the marketing campaign -- Building propensity segments -- Designing a deployment procedure and disseminating the results -- Using classification models in direct marketing campaigns -- Acquisition modeling -- Pilot campaign -- Profiling of high-value customers -- Cross-selling modeling -- Pilot campaign -- Product uptake -- Profiling of owners -- Offer optimization with next best product campaigns -- Deep-selling modeling -- Pilot campaign -- Usage increase -- Profiling of customers with heavy product usage -- Up-selling modeling -- Pilot campaign -- Product upgrade -- Profiling of "premium" product owners -- Voluntary churn modeling -- Summary of what we've learned so far: it's not about the tool or the modeling algorithm.
|
|
|
:
|
It's about the methodology and the design of the process -- Behavioral segmentation methodology -- An introduction to customer segmentation -- An overview of the behavioral segmentation methodology -- Business understanding and design of the segmentation process -- Definition of the business objective -- Design of the modeling process -- Selecting the segmentation population -- Selection of the appropriate segmentation criteria -- Determining the segmentation level -- Selecting the observation window -- Data understanding, preparation, and enrichment -- Investigation of data sources -- Selecting the data to be used -- Data integration and aggregation -- Data exploration, validation, and cleaning -- Data transformations and enrichment -- Input set reduction -- Identification of the segments with cluster modeling -- Evaluation and profiling of the revealed segments -- "Technical" evaluation of the clustering solution -- Profiling of the revealed segments -- Using marketing research information to evaluate the clusters and enrich their profiles -- Selecting the optimal cluster solution and labeling the segments -- Deployment of the segmentation solution, design and delivery of differentiated strategies -- Building the customer scoring model for updating the segments -- Building a Decision Tree for scoring: fine-tuning the segments -- Distribution of the segmentation information -- Design and delivery of differentiated strategies -- Summary -- The Algorithms -- Classification algorithms -- Data mining algorithms for classification -- An overview of Decision Trees -- The main steps of Decision Tree algorithms -- Handling of predictors by Decision Tree models -- Using terminating criteria to prevent trivial tree growing -- Tree pruning -- CART, C5.0/C4.5, and CHATD and their attribute selection measures -- The Gini index used by CART -- The Information Gain Ratio index used by C5.0/C4.5 -- The chi-square test used by CHAID -- Bayesian networks -- Naïve Bayesian networks -- Bayesian belief networks -- Support vector machines -- Linearly separable data -- Linearly inseparable data -- Summary -- Segmentation algorithms -- Segmenting customers with data mining algorithms -- Principal components analysis -- How many components to extract? -- The eigenvalue (or latent root) criterion -- The percentage of variance criterion -- The scree test criterion -- The interpretability and business meaning of the components -- What is the meaning of each component? -- Moving along with the component scores -- Clustering algorithms -- Clustering with K-means -- Clustering with TwoStep -- Summary -- The Case Studies -- A voluntary churn propensity model for credit card holders -- The business objective -- The mining approach -- Designing the churn propensity model process -- Selecting the data sources and the predictors -- Modeling population and level of data -- Target population and churn definition -- Time periods and historical information required -- The data dictionary -- The data preparation procedure -- From cards to customers: aggregating card-level data -- Enriching customer data -- Defining the modeling population and the target field -- Derived fields: the final data dictionary -- The modeling procedure -- Applying a Split (Holdout) validation: splitting the modeling dataset for evaluation purposes -- Balancing the distribution of the target field -- Setting the role of the fields in the model -- Training the churn model -- Understanding and evaluating the models -- Model deployment: using churn propensities to target the retention campaign -- The voluntary churn model revisited using RapidMiner -- Loading the data and setting the roles of the attributes -- Applying a Split (Holdout) validation and adjusting the imbalance of the target field's distribution -- Developing a Naïve Bayes model for identifying potential churners -- Evaluating the performance of the model and deploying it to calculate churn propensities -- Developing the churn model with Data Mining for Excel -- Building the model using the Classify Wizard -- Selecting the classification algorithm and its parameters -- Applying a Split (Holdout) validation -- Browsing the Decision Tree model -- Validation of the model performance -- Model deployment -- Summary -- Value segmentation and cross-selling in retail -- The business background and objective -- An outline of the data preparation procedure -- The data dictionary -- The data preparation procedure -- Pivoting and aggregating transactional data at a customer level -- Enriching customer data and building the customer signature -- The data dictionary of the modeling file -- Value segmentation -- Grouping customers according to their value -- Value segments: exploration and marketing usage -- The recency, frequency, and monetary (RFM) analysis -- RFM basics -- The RFM cell segmentation procedure -- Setting up a cross-selling model -- The mining approach -- Designing the cross-selling model process -- The data and the predictors -- Modeling population and level of data -- Target population and definition of target attribute -- Time periods and historical information required -- The modeling procedure -- Preparing the test campaign and loading the campaign responses for modeling -- Applying a Split (Holdout) validation: splitting the modeling dataset for evaluation purposes -- Setting the roles of the attributes -- Training the cross-sell model -- Browsing the model results and assessing the predictive accuracy of the classifiers -- Deploying the model and preparing the cross-selling campaign list -- The retail case study using RapidMiner -- Value segmentation and RFM cells analysis -- Developing the cross-selling model -- Applying a Split (Holdout) validation -- Developing a Decision Tree model with Bagging -- Evaluating the performance of the model -- Deploying the model and scoring customers -- Building the cross-selling model with Data Mining for Excel -- Using the Classify Wizard to develop the model -- Selecting a classification algorithm and setting the parameters -- Applying a Split (Holdout) validation -- Browsing the Decision Tree model -- Validation of the model performance -- Model deployment -- Summary -- Segmentation application in telecommunications -- Mobile telephony: the business background and objective -- The segmentation procedure -- Selecting the segmentation population: the mobile telephony core segments -- Deciding the segmentation level -- Selecting the segmentation dimensions -- Time frames and historical information analyzed -- The data preparation procedure -- The data dictionary and the segmentation fields -- The modeling procedure -- Preparing data for clustering: combining fields into data components -- Identifying the segments with a cluster model -- Profiling and understanding the clusters -- Segmentation deployment -- Segmentation using RapidMiner and K-means cluster -- Clustering with the K-means algorithm -- Summary
|
| Subject
|
:
|
Customer relations-- Management-- Data processing.
|
| Subject
|
:
|
Data mining.
|
| Dewey Classification
|
:
|
658.8/12
|
| LC Classification
|
:
|
HF5415.5
|
| Added Entry
|
:
|
Ohio Library and Information Network.
|
| |