PPT Slide
- Do not see the whole data set or hold it in computer memory.
- Difficult to apply current software of exploratory data analysis which are often based on matrix computation.
- US Voting data set: 435 inst X 16 att. --> 150,000 patterns in terms of the classical view on concepts (by intent and extent).
Efficiency and Scalability of DM alg.
Massive Data Sets and High Dimensionality