Week 1 - CRISP-DM Steps Flashcards
In CRISP-DM, what part of Data Understanding is this?
check quantity of data, data format, consistent coding schemes
Describe data
In CRISP-DM, what part of Data Understanding is this?
visualize data, identify relationships among data, query data etc.
Explore data
In CRISP-DM, what part of Data Understanding is this?
how clean is the data? Any noise? Could refer to missing data, errors, inconsistent representation
Verify data quality
In CRISP-DM, what part of Data Preparation is this?
determine which data will be used and document reasons for inclusion/exclusion
Select data
In CRISP-DM, what part of Data Preparation is this?
solve all quality issues
Clean data
In CRISP-DM, what part of Data Preparation is this?
derive new attributes
Construct data
In CRISP-DM, what part of Data Preparation is this?
create new datasets by combining data from multiple sources
Integrate data
In CRISP-DM, what part of Data Preparation is this?
re-format data as necessary (discretization etc.)
Format data
In CRISP-DM, what part of Modeling is this?
determine which algorithms to try
SMT
Select modeling technique
In CRISP-DM, what part of Modeling is this?
how to split data for training, testing, validation, etc.
Generate test design
In CRISP-DM, what part of Modeling is this?
build decided models
Build model
In CRISP-DM, what part of Modeling is this?
check generated models, apply domain knowledge to interpret the results
Assess model
In CRISP-DM, what part of Evaluation is this?
do the models meets the business criteria? Which ones should we approve?
Evaluate results
In CRISP-DM, what part of Evaluation is this?
Review the work
Review process
In CRISP-DM, what part of Evaluation is this?
Determine whether to proceed or iterate further
Determine next steps
