Data Mining for Business Decisions Notes
Subject Overview
The first Management Information Systems group elective, covering data warehousing and business intelligence fundamentals, the data mining process and its various kinds, classification/
Subject at a glance
4
Units
28
Topics
4
Credits
4
Infographics
Unit-wise Syllabus
4 units — open any unit for detailed topic-wise notes, diagrams and important exam questions
- Strategic information needs
- operational vs informational data stores
- data warehouse definition/characteristics/role/structure
- introduction to business intelligence
- OLAP operations
- data marts
- dimensional modeling and ETL process
Roll-up
Summarise: city to region
Drill-down
More detail: year to month
Slice
Fix one dimension
Dice
Select a sub-cube
Pivot
Rotate the view
Study guide
This unit covers strategic information needs, warehouses, BI, OLAP, data marts, dimensional modelling and ETL. Draw a data warehouse architecture.
- Data mining process and functionalities
- text/web/pattern/sequence/context-based mining
- data visualization
- predictive vs descriptive data mining
- need for data analytics in business intelligence
- 1
Data selection
- 2
Pre-processing and cleaning
- 3
Transformation
- 4
Data mining
Find patterns
- 5
Interpretation and evaluation
- 6
Knowledge for decisions
Study guide
You study the data mining process and functionalities, text, web, pattern and sequence mining, visualisation and predictive versus descriptive mining. Match business problems to mining tasks.
- Regression and correlation
- decision trees
- clustering
- neural networks
- market basket analysis and association rules
- genetic algorithms and link analysis
- Support Vector Machine
- Bayesian classification
- k-Nearest Neighbour
- case-based reasoning
- fuzzy set approach
Decision tree
If-then splits
Easy to explain
Naive Bayes
Probabilities with Bayes' theorem
Fast, works with little data
k-NN
Nearest neighbours vote
No training phase
SVM
Maximum-margin boundary
Accurate in high dimensions
Neural network
Layers of neurons
Complex patterns
Study guide
You study regression, decision trees, clustering, neural networks, association rules, genetic algorithms, link analysis, SVM, Bayesian classification, k-NN and case-based reasoning. Compare methods by use and limitation.
- Linear and multiple regression prediction
- cluster analysis data types (interval-scaled, binary, nominal, ordinal, ratio-scaled)
- partitioning methods (K-Means, K-Medoids)
- hierarchical agglomerative methods
- DBSCAN
K-means
Assign to nearest mean, update means
Fast; needs k
K-medoids
Uses actual points as centres
Robust to outliers
Hierarchical agglomerative
Merge closest clusters step by step
Gives a dendrogram
DBSCAN
Grows dense regions
Finds odd shapes and noise
Study guide
You study regression prediction, cluster analysis of different data types, K-means, K-medoids, hierarchical methods and DBSCAN. Run a clustering on customer data.
Why Data Mining for Business Decisions Matters
Data mining turns data into business decisions, supporting analyst and data science careers.
How to Prepare for Data Mining for Business Decisions
- Apply each technique to a real data set with a tool.
- Learn the strengths and weaknesses of each method.
- Always validate models on new data.
Data Mining for Business Decisions: Frequently Asked Questions
What is market basket analysis?
It finds items that are often bought together by mining association rules from transaction data.