1 of 11

LECTURE 3

dr. Jamolbek Mattiev

The importance of data generation and big data

2 of 11

Knowledge discovery in data is the�non-trivial process of identifying:

      • valid
      • novel
      • potentially useful
      • and ultimately understandable patterns in data.

From:�Advances in Knowledge Discovery and Data Mining, Fayyad, Piatetsky-Shapiro, Smyth, and Uthurusamy, (Chapter 1),�AAAI/MIT Press 1996.

2 / 38

Data flood 🡪 application examples 🡪 ●●●●● 🡪 data mining tasks 🡪 summary

Definition of�“knowledge discovery”

3 of 11

Visualization

3 / 38

Data flood 🡪 application examples 🡪 ●●●●● 🡪 data mining tasks 🡪 summary

Related fields

Machine learning

Statistics

Databases

Data mining and/or�Knowledge discovery

4 of 11

Statistics, machine learning�and data mining

  • Statistics:
      • more theory-based
      • more focused on testing hypotheses
  • Machine learning:
      • more heuristic
      • focused on improving performance of a learning agent
      • also looks at real-time learning and robotics – areas not part�of data mining
  • Data mining and/or Knowledge discovery in data:
      • integrates theory and heuristics
      • focus on the entire process of knowledge discovery, including data “cleaning”, learning, integration and visualization of results
  • Distinctions are “fuzzy”.

4 / 38

Data flood 🡪 application examples 🡪 ●●●●● 🡪 data mining tasks 🡪 summary

5 of 11

Knowledge discovery process�flow – according to CRISP-DM

Monitoring

See also:

https://en.wikipedia.org/wiki/Cross_Industry_Standard_Process_for_Data_Mining

for more detailed information

5 / 38

Data flood 🡪 application examples 🡪 ●●●●● 🡪 data mining tasks 🡪 summary

6 of 11

Data Fishing, Data Dredging: 1960 –

    • used by statisticians (considered as a bad name);

Data Mining: 1990 –

    • used DB, business;
    • in 2003 – bad image because of TIA;

Knowledge Discovery in Databases: 1989 –

    • used by AI, machine learning community;

also:�Data Archaeology, Information Harvesting, Information Discovery,�Knowledge Extraction ...

Currently:

Data Mining and Knowledge Discovery�are used interchangeably (as synonyms).

6 / 38

Data flood 🡪 application examples 🡪 ●●●●● 🡪 data mining tasks 🡪 summary

Historical note: �the many names of data mining

7 of 11

Lesson outline

  • Introduction: data flood
  • Data mining application examples
  • Data mining & knowledge discovery
  • Data mining tasks

7 / 38

Data flood 🡪 application examples 🡪 terminology 🡪 data mining tasks 🡪 summary

8 of 11

Classification: predicting an item class

Clustering: finding clusters in data

Associations: e.g. A & B & C occur frequently

Visualization: to facilitate human discovery

Summarization: describing a group

Deviation detection: finding changes

Regression/estimation: predicting a continuous value

Link analysis: finding relationships

8 / 38

Data flood 🡪 application examples 🡪 terminology 🡪 ●●● 🡪 summary

Major data mining tasks

9 of 11

Data mining tasks:�prediction (classification)

Learn a method for predicting the instance class from pre-labeled (classified) instances

Many approaches:�

statistics,

decision trees,�neural networks, ...

9 / 38

Data flood 🡪 application examples 🡪 terminology 🡪 ●●● 🡪 summary

10 of 11

Data mining tasks: clustering

Find “natural” grouping of instances given un-labeled data

10 / 38

Data flood 🡪 application examples 🡪 terminology 🡪 ●●● 🡪 summary

11 of 11

Summary

  • Technology trends lead to data flood
    • data mining is needed to make sense of data;
  • Data Mining has many applications,�successful and not;
  • Knowledge discovery is a process;
  • Data mining tasks
    • classification, clustering, …

11 / 38

Data flood 🡪 application examples 🡪 terminology 🡪 data mining tasks 🡪 summary