- acquisition modeling
- pilot campaign
- profiling of high-value customers
- AIC see Akaike Information Criterion
- Akaike Information Criterion
- association (affinity) models
- apriori and FP-growth algorithms
- market basket analysis
- rule’s confidence
- sequence algorithms
- Bagging (Bootstrap aggregation)
- Decision Tree model
- in IBM SPSS Modeler
- in RapidMiner
- balancing approach
- balance factor
- cross-selling
- in Data Mining for Excel
- disproportionate stratified sampling
- in IBM SPSS Modeler
- oversampling
- in RapidMiner
- Bayesian belief networks
- CPT see conditional probability table
- IBM SPSS Modeler
- Microsoft Naïve Bayes
- Naive Bayes models
- parent variables
- RapidMiner Naïve Bayes
- Tree Augmented Naive Bayes models
- Bayesian Information Criterion
- Bayesian networks
- behavioral segmentation methodology
- business objective, definition
- cluster modeling see cluster modeling, identification of segments
- CRISP DM methodology, phases
- customer segmentation
- data exploration and validation
- data integration and aggregation
- data transformations and enrichment
- deployment of segmentation solution
- customer scoring model
- design and deliver of differentiated strategies
- distribution of segmentation information
- input set reduction
- investigation of data sources
- modeling process design
- revealed segments
- marketing research information
- optimal cluster solution and labeling segments
- profiling of
- technical evaluation of clustering solution
- BIC see Bayesian Information Criterion
- BIRCH algorithm
- boosting
- Adaptive Boosting (AdaBoost)
- in IBM SPSS Modeler
- in RapidMiner
- Bootstrap validation method
- Call Detail Records
- candidate model
- CHAID see CHAID model
- CHURN model
- ensemble model
- Gains chart
- candidate models, performance measures
- CART see classification and regression trees
- CDRs see Call Detail Records
- C5.0/C4.5
- information gain ratio index
- of child nodes
- Information or Entropy of node
- 2-level C5.0 Decision Tree
- of root node
- split on profession
- CHAID model
- and C5.0
- chi-square test see chi-square test, CHAID
- churn scores
- and ensemble model
- handling of predictors
- parameter
- right branches
- in rules format
- spending frequency
- in tree format
- chi-square test, CHAID
- cross-tabulation of target
- Decision Tree algorithm
- IBM SPSS Modeler CART
- advanced options
- basics options
- parameters
- stopping criteria
- IBM SPSS Modeler C5.0 Decision Trees
- IBM SPSS Modeler CHAID Decision Trees
- advanced options
- 2-level
- parameters
- stopping rules
- independence hypothesis
- Microsoft Decision Trees parameters
- Pearson chi-square statistic
- predictors for split
- p-value or observed significance level
- RapidMiner Decision Trees
- parameters
- recursive partitioning
- classification algorithms
- Bayesian belief networks
- Bayesian networks
- chi-square test, CHAID
- data mining algorithms
- Decision Tree models
- Gini index, CART
- information gain ratio index, C5.0/C4.5
- Naive Bayesian networks
- support vector machines
- classification and regression trees
- Gini index
- handling of predictors
- classification modeling methodology
- acquisition modeling
- business understanding and process design
- combining models
- CRISP-DM phases
- cross-selling modeling
- in Data Mining for Excel
- data preparation and enrichment
- deep-selling modeling
- direct marketing campaigns
- form of supervised modeling
- in IBM SPSS Modeler
- likelihood of prediction
- meta-modeling
- model deployment
- model evaluation
- product campaigns, optimization
- in RapidMiner
- up-selling modeling
- voluntary churn modeling
- classification or propensity models
- Bayesian networks
- decision rules
- Decision Trees
- logistic regression
- neural networks
- support vector machine
- class weighting
- class-imbalanced modeling file
- in IBM SPSS Modeler
- in RapidMiner
- clustering algorithms
- cluster modeling, identification of segments
- agreement level
- in Data Mining for Excel
- in IBM SPSS Modeler
- profiling
- cluster centroid
- cluster separation
- in Data Mining for Excel
- effective marketing strategies, development
- in IBM SPSS Modeler
- in RapidMiner
- table of cluster centers
- in RapidMiner
- revealed segments
- cohesion of clusters
- descriptive statistics and technical measures
- in IBM SPSS Modeler
- in RapidMiner
- separation of clusters
- cluster models
- agglomerative or hierarchical
- data preparation
- data miners of organization
- data reduction technique
- explained variance
- interpretation results
- principal components analysis model
- rotated component matrix
- expectation maximization clustering
- identifying segments
- cluster models, comparison
- parameter settings
- revealed clusters, distribution
- Silhouette measure
- K-means
- K-medoids
- Kohonen network/self-organizing map
- profiling
- behavioral profile
- distribution of factors for Cluster
- profiling chart
- profiling of clusters
- structures
- table of centroids
- RapidMiner process
- TwoStep cluster
- conditional probability table
- of gender input attribute
- probabilities of output
- of profession input attribute
- of SMS calls input attribute
- of voice calls input attribute
- confusion matrix and accuracy measures
- in Data Mining for Excel
- error rate
- F-measure
- in IBM SPSS Modeler
- misclassification or coincidence matrix
- Performance operator
- Precision measure
- in RapidMiner
- Recall measure
- ROC curve
- sensitivity and specificity
- CPT see conditional probability table
- CRISP-DM see Cross Industry Standard Process for Data Mining process model
- CRM see customer relationship management
- Cross Industry Standard Process for Data Mining process model
- business understanding
- data preparation
- data understanding
- deployment
- evaluation
- modeling
- phases
- Cross or n-fold validation method
- in Data Mining for Excel
- modeling dataset
- n iterations
- in RapidMiner
- cross-selling modeling
- browsing the model
- C5.0 model
- CHAID model
- CPTs
- FLAG_GROCERY attribute
- Gains chart
- gains chart of ensemble model
- performance metrics for individual models
- response rate
- ROI chart
- TAN
- campaign list
- Modeler deployment stream
- scored customers, estimated fields
- Data Mining for Excel
- Accuracy Chart wizard
- campaign response
- classification algorithm and parameters
- Classification Matrix wizard
- Classify Wizard
- confusion matrix
- cumulative percentage of responders
- Decision Tree model
- dependency network of BDE tree
- Gains charts for two Decision Tree models
- model deployment
- Split (Holdout) validation
- validation dataset
- validation of model performance
- development of
- mining approach
- modeling procedure
- IBM SPSS Modeler procedure
- setting roles of attributes
- Split (Holdout) validation
- test and loading campaign responses
- training
- parameters
- pilot campaign
- product uptake
- profiling of owners
- customer relationship management
- customer development
- customer satisfaction
- data mining
- customer scoring model
- in Data Mining for Excel
- Decision Tree
- deployment procedure
- in IBM SPSS Modeler
- in RapidMiner
- customer segmentation
- behavioral
- definition
- loyalty based
- needs/attitudinal
- propensity based
- sociodemographical
- value based
- customers grouping, value segmentation
- binning node
- binning procedure
- Data Audit node
- high-value customers
- investigation of characteristics
- medium-and low-value customers
- quantiles
- regrouping quantiles into value segments
- RFM segmentation
- segmentation bands selected by retailer
- and total purchase amount
- data dictionary
- card level, voluntary churn model
- closing date and reason
- demographical input data
- of modeling file
- time periods, model training phase
- transactional input data
- usage attributes
- data enrichment
- customer signature
- data reduction algorithm
- feature selection
- informative KPIs
- naming of attributes
- data exploration
- assessment of data quality
- categorical attributes
- continuous (range) attributes
- tool of IBM SPSS Modeler
- data integration and aggregation
- data management procedure, churn model
- from cards to customers
- cards’ closing dates
- filtering out cards
- flagging cards
- IBM SPSS Modeler node
- individual card records, grouping
- initial card usage data
- open at end, observation period
- total number of transactions, calculation
- enrichment, customer data
- balances
- deltas, spending
- limit ratios
- monthly average number, transactions
- spending amount, monthly average
- spending frequency
- spending recency
- tenure, customer
- trends, card ownership
- modeling population and target field
- defined
- latency period
- REFERENCE DATE
- in scoring phase
- selection
- short-term churners
- data mining
- algorithms
- CRISP-DM
- CRM strategy
- customer life cycle management
- customer segmentation
- datamart
- direct marketing campaigns
- market basket and sequence analysis
- marketing reference table
- personalized customer handling
- required data per industry
- supervised models
- unsupervised models
- Data Mining for Excel
- balancing approach
- churn model
- accuracy and error rate
- approaches
- churners and nonchurners, cumulative distribution
- classification algorithm
- Classify Wizard
- confusion matrix
- Decision Tree model
- mining structure, storing
- Query wizard
- scored customers and model derived estimates
- Split (Holdout) validation
- validation, performance
- classification modeling methodology
- cluster modeling
- confusion matrix and accuracy measures
- Cross or n-fold validation method
- in cross-selling model see cross-selling modeling
- customer scoring model
- Gains/Response/Lift charts
- K-means
- Naive Bayesian networks
- Classify Wizard
- Dependency network
- receiver operating characteristic curves
- scoring customers
- Split (Holdout) validation method
- data preparation procedure
- aggregating at customer level
- aggregating at transaction level
- adding demographics using merge node
- at customer level
- invoice level
- categorizing transactions into time zones
- classification modeling, tasks
- customer level, usage aspects
- data exploration and validation
- data integration and aggregation
- data transformations and enrichment
- enrich customer information
- average basket size
- basket diversity
- customer tenure
- deriving new fields
- flags of product groups
- frequency
- monetary value
- monthly average purchase amount
- ratio of transactions
- recency
- relative spending
- IBM SPSS Modeler Derive nodes
- imbalanced outcomes
- initial transactional data
- investigation of data sources
- Modeler datetime_weekday() function
- pivoting transactional data
- payment type
- series of Restructure nodes
- selecting data sources
- validation techniques
- data transformations
- event outcome period
- label or target attribute
- optimal discretization or binning
- data validation process
- Decision Tree model
- algorithms
- attribute selection method
- classes of target attribute
- decision rules
- handle predictors
- root node and
- user-specified terminating criteria
- with bagging
- Bagging (Bootstrap aggregation)
- churn model
- classification algorithms
- classification or propensity models
- cross-selling model
- customer scoring model
- “divide-and-conquer” procedure
- handling of predictors
- binary splits
- C5.0/C4.5 and CHAID
- Classification and Regression Trees
- IBM SPSS Modeler C5.0
- IBM SPSS Modeler CHAID
- Microsoft Decision Trees parameters
- modeling dataset
- Random Forests
- RapidMiner
- supervised segmentation
- tree pruning
- using terminating criteria
- prepruning or forward pruning
- split-and-grow procedure
- deep-selling modeling
- pilot campaign
- profiling of customers
- usage increase
- dimensionality reduction models
- direct marketing campaigns
- eigenvalue (or latent root) criterion
- of components
- percentage of variance/information
- z-score method
- EM see Expectation Maximization clustering
- estimation (regression) models
- linear or nonlinear functions
- ordinary least squares regression
- Expectation Maximization clustering
- factor analysis
- feature selection (field screening)
- Gains/Response/Lift charts
- binary classification problem
- churn model
- creation of charts
- cumulative Lift or Index chart
- in Data Mining for Excel
- Evaluation Modeler node
- Gains Chart
- in IBM SPSS Modeler
- Kolmogorov–Smirnov statistic
- performance measures
- in RapidMiner
- Response chart
- Gini index, CART
- child nodes
- distribution of target classes
- IBM SPSS Modeler CART
- purity improvement
- of root node
- splits and predictors
- voice and SMS usage
- IBM SPSS Modeler
- Bagging (Bootstrap aggregation)
- balancing approach
- Bayesian belief networks
- boosting
- CART
- C5.0 Decision Trees
- CHAID Decision Trees
- classification modeling methodology
- class weighting
- cluster modeling
- confusion matrix and accuracy measures
- cross-selling modeling
- customer scoring model
- data exploration
- derive nodes
- Gains/Response/Lift charts
- K-means
- mobile telephony
- principal components analysis
- profiling
- receiver operating characteristic curves
- revealed segments
- scoring customers
- Split (Holdout) validation method
- stream (procedure), churn modeling
- Auto-Classifier node
- Balance node
- derived fields/candidate predictors
- description
- initial and balanced distribution
- modeling steps
- parameters
- Split (Holdout) validation and Partition node
- Three Decision Tree and SVM model
- Type node, setting
- undersampling
- support vector machines
- Tree Augmented Naïve Bayesian network
- TwoStep
- ICA see independent component analysis
- imbalanced outcome distribution
- balancing
- class weighting
- independent component analysis
- K-means, clustering algorithms
- Bayesian Information Criterion
- centroid-based partitioning technique
- centroids of identified clusters
- in Data Mining for Excel
- Euclidean distance
- IBM SPSS Modeler
- K-medoids
- Modeler’s
- RapidMiner K-means and K-medoids cluster
- K-medoids
- Kohonen network/self-organizing map
- Kolmogorov–Smirnov statistic
- KS see Kolmogorov–Smirnov statistic
- market basket analysis
- marketing reference table
- aggregations/group by
- deltas
- derive
- filtering of records
- flag fields
- joins
- ratios (proportions)
- restructure/pivoting
- sums/averages
- maximum marginal hyperplane
- meta-modeling or ensemble modeling
- mining approach
- cross-selling model
- data and predictors
- modeling population and level of data
- target population and attribute
- time periods and historical information
- and data model
- cross-selling campaign
- pilot campaign approach
- product possession approach
- product uptake approach
- voluntary churn propensity model
- data sources and predictors, selection
- modeling population and data level
- target population and churn definition
- time periods and historical information
- mining datamart
- marketing reference table
- of mobile telephony operator
- of retail banking
- of retailers
- MMH see maximum marginal hyperplane
- mobile telephony
- behavioral segmentation
- Call Detail Records
- clustering, IBM SPSS Modeler procedure
- core segments
- high-level quality services
- modeling steps
- organization’s mining datamart
- segmentation fields
- SMS and MMS messages
- model deployment
- churn propensities
- defined, churn
- ensemble model
- propensity-based segmentation
- scored customers, sample
- voluntary churn model scoring procedure
- $XF-CHURN field
- direct marketing campaigns
- procedure and results
- scoring customers, marketing campaign
- model evaluation procedure
- accuracy measures and confusion matrices
- gains, response, and lift charts
- precampaign model validation
- profit/ROI charts
- RapidMiner modeling process
- confidence(T) field
- model deployment
- ROC curve
- Split Validation operator
- ROC curve
- test-control groups
- modeling process design
- behavioral segmentation methodology
- determining segmentation level
- selecting observation window
- selecting segmentation population
- selection of appropriate segmentation criteria
- classification modeling methodology
- defining modeling population
- determining modeling (analysis) level
- target event and population
- time frames
- Naive Bayesian networks
- Apply Model operator
- Attribute Characteristics
- Attribute Profiles output
- Data Mining for Excel
- conditional probability
- normal distribution assumption
- RapidMiner process
- Naïve Bayes model
- Create Lift Chart operator
- with Laplace correction
- ROC curve
- PCA see principal components analysis
- principal components analysis,
- clustering
- component scores
- IBM SPSS Modeler
- RapidMiner
- components to extract
- behavioral fields
- eigenvalue (or latent root) criterion
- interpretability and business meaning
- pairwise correlation coefficients
- percentage of variance criterion
- scree test criterion
- data reduction
- linear correlation between continuous measures
- meaning of component
- interpretation process
- Modeler
- in RapidMiner
- rotation techniques
- model
- reduction of dimensionality
- Random Forests
- Decision Tree models
- in RapidMiner
- RapidMiner modeling process
- Attribute operator
- Bagging (Bootstrap aggregation)
- balancing approach
- boosting
- chi-square test, CHAID
- classification modeling methodology
- class weighting
- cluster modeling
- confusion matrix
- Cross or n-fold validation method
- cross-selling model
- customer scoring model
- Decision Tree model with bagging
- Gains/Response/Lift charts
- K-means and K-medoids cluster
- model evaluation procedure
- Naïve Bayes model
- predictors
- principal components analysis
- profiling
- Random Forests
- receiver operating characteristic curves
- retail case study see retail case study, RapidMiner
- revealed segments
- scoring customers
- Set Role settings
- Split (Holdout) validation method
- SVM models
- value segmentation and RFM cells analysis
- receiver operating characteristic curves
- area under the curve measure
- confusion matrix and accuracy measures
- Gains chart
- Gini index
- in IBM SPSS Modeler
- model evaluation
- Naïve Bayes model
- performance of model
- Profit/ROI charts
- customers
- in Data Mining for Excel
- in IBM SPSS Modeler
- marketers
- in RapidMiner
- sensitivity
- recency, frequency, and monetary analysis
- cell segmentation procedure
- data preparation phase
- distribution
- clustering model
- components
- cross-selling models
- grouping (binning) of customers
- indicators, construction of
- monitoring consuming behaviors
- quintiles, grouping customers
- in retail industry
- scatter plot
- regression models see estimation (regression) models
- retail case study, RapidMiner
- cross-selling model
- Decision Tree model with bagging
- bagging operator
- parameter settings
- in tree format
- performance of model
- confusion matrix
- ROC curve
- scoring customers
- model deployment process
- prediction fields
- Split (Holdout) validation
- value segmentation and RFM cells analysis
- RFM see recency, frequency, and monetary analysis
- ROC see receiver operating characteristic curves
- scoring customers, marketing campaign
- binary classification problems
- Create Threshold operator
- in Data Mining for Excel
- Gains/Profit/ROC charts and tables
- in IBM SPSS Modeler
- probabilistic classifiers
- propensity segmentation
- in RapidMiner
- scree test criterion
- segmentation algorithms
- clustering algorithms
- with K-means
- with TwoStep
- sequence algorithms
- Split (Holdout) validation method
- churn modeling
- cross-selling modeling
- in Data Mining for Excel
- distribution of target attribute
- in IBM SPSS Modeler
- model training
- performance metrics
- random sampling
- in RapidMiner
- retail case study
- supervised modeling
- classification or propensity models
- estimation (regression) models
- feature selection (field screening)
- support vector machines
- linearly inseparable data
- IBM SPSS Modeler
- Kernel functions
- Polynomial transformation
- RapidMiner SVM models
- linearly separable data
- linear discriminant function
- maximum marginal hyperplane
- separating hyperplane
- nonlinear mappings for classification
- SVM see support vector machines
- TAN see Tree Augmented Naïve Bayesian network
- telecommunications, segmentation application
- data dictionary and segmentation fields
- data preparation procedure
- mobile telephony
- modeling procedure
- identifying segments with cluster model
- preparing data for clustering
- profiling and understanding clusters
- segmentation deployment
- segmentation procedure
- deciding level
- dimensions
- population, mobile telephony core segments
- time frames and historical information analyzed
- using RapidMiner and K-means cluster
- Cluster Distance Performance operator
- clustering with K-means algorithm
- Euclidean distance
- K-means parameter settings
- mobile telephony segments
- PCA algorithm
- profile of clusters
- variance/information, by components
- test-control groups
- direct marketing campaign
- Model Holdout group
- Random Holdout group
- recorded response rate, cross-selling campaign
- time frames
- in churn model
- customer profiles
- event outcome period
- latency period
- multiple time frames
- observation (historical) period
- potential voluntary churners, identification of
- validation phase
- Tree Augmented Naïve Bayesian network,
- IBM SPSS Modeler
- structure
- training dataset
- TwoStep cluster
- Akaike Information Criterion
- Bayesian Information Criterion
- IBM SPSS Modeler
- preclusters
- unsupervised models
- association (affinity) and sequence models
- cluster models
- dimensionality reduction models
- record screening models
- up-selling modeling
- pilot campaign
- product upgrade
- profiling of premium product owners
- validation techniques
- Bootstrap validation
- Cross or n-fold validation
- Split (Holdout) validation method
- value segmentation
- and cross-selling in retail
- data dictionary
- data preparation procedure
- exploration and marketing usage
- grouping customers see customers grouping, value segmentation
- mining approach
- modeling procedure
- predictive accuracy of classifiers
- recency, frequency, and monetary analysis
- retail case study using RapidMiner
- transactional data
- and RFM cells analysis
- discretization of numeric fields
- Map operator, value segments
- MONETARY attribute
- relevant binned attributes
- total purchase amount
..................Content has been hidden....................
You can't read the all page of ebook, please click
here login for view all page.