Data mining needs
Need: one from a lot of, incomplete, noisy, fuzzy and random data (usually exist in large database or data warehouse) to find and extract implicit, potential, valuable information or knowledge (or rule) model of process
Objective: is the relationship between help analyst to find data, found neglected elements, and these information for predicting trends and decision-making is very useful,
Example: the failure correlation analysis problems in network management system, find fault source,
Can solve the problem from the source,
Class review: class notes 1, ETL, data mining process, the decision tree
Factors: feedback, learn
Basic requirements: artificial intelligence incremental learning,
Question: big data (data warehouse), the more the better, for the AI systems have help?
- correlation analysis: is the discovery of a large amount of data of interesting correlation between itemsets,
- based on original data, the information such as association rules and credibility, and search for potential relationship from a large amount of data,
- correlation can be divided into simple associations, temporal association, causal relation,
Mind wandering: PDI algorithm has feedback, adjustment, in artificial intelligence, in the broadest sense of the
