Lecture 5: CRISP-DM Flashcards
1
Q
Consultancy segments
A
- Strategy
- Management
- Operations
- Human Resources
- Financial Advisory
- Technology (Data Analytics)
2
Q
Client problems can be:
A
- Strategic
- Tactical
- Operational
3
Q
Data Science Consulting Firms
A
- BCG
- McKinsey
- Bain
4
Q
CRISP-DM
A
Cross-Industry Standard Process for Data Mining
5
Q
CRISP-DM Phases
A
- Business Understanding
- Data Understanding
- Data Preparation
- Modeling
- Evaluation
- Deployment
- Back and forward movement
- Outcome determines which phase/task next
6
Q
Business Understanding
A
- Understand project objective & requirements
- Business perspective
7
Q
Data Understanding
A
- Data Collection
- Familiarity with data
- Identify quality issues
- First insights
8
Q
Data Preparation
A
- Make data ready for modelling
- Select variables
- Transform data
- Clean data
- Create derived attributes
9
Q
Modeling
A
- Select and apply modelling techniques
- Parameter tuning
- Datât should suit modelling technique (step back)
10
Q
Evaluation
A
- Check alignment model with business objectives
- Decide about use of model
11
Q
Deployment
A
- Projects do not end with the creation of models
- Deployment depends on project goals
