In data interesting patterns are hidden - information about succesful medicines, profitable customers or indications of potential terrorist behavior. Data mining can help to discover useful and actionable knowledge. What kind of data mining tasks and technologies exist and how can these be applied? These and other topics will be discussed in the LAPP-Top class on data mining and knowledge discovery.

Links:

• Sheets first lecture, sheets second lecture
• Assignments first & second lecture
• WEKA, open source data mining tool used in the assignments, and the data sets used.
• Data mining is not just rocket science, people use it every day!
• KDNuggets, the premier data mining website.
• Classifier applet, a webbased demo of the behavior of various classifiers in pattern space (auteur Ran El-Yaniv)
• Recommender Examples: Amazon, BOL, Proxis, Romandvies@bibliotheek.nl, Centrale Discotheek Rotterdam, GNOD (Music, Books, Movies, People), TiVo (Digital Videorecorder), Internet Movie Database, Reel.com
• Recommender Reseach Groups & Papers: Grouplens, E-Commerce Recommendation Applications (Schafer et al), Amazon.com recommendations (Linden et al)
• As part of their lab work the students had one hour to mine a data set of their choice and present the results to the group - they were allowed to be creative about the purpose and background of the data set. Presentations: Recognizing Vehicles at Toll Gates, Weather Prediction for Sports Games, Speech Recogition: Vowels, What to Vote, Identifying Glass Types for Criminal Analysis
• LAPP-top is part of the Pre University College @ Leiden university; it allows high school students (5 and 6VWO) to follow university classes