This thesis presents an efficient method for mining quantitative association rules for finitely discrete quantitative data. Our main contribution is a new view on what constitutes an unexpected relation, a view that is believed to be more accurate than previous models. To ensure the statistical significance of the results, a non-parametric exact test is developed that does not require data to be normally distributed, works with sample sizes as low as a single sample and has a run time proportional to the sample size. In addition to mining rules, we show how memory-based reasoning can be used to learn and to predict the users’ interests in rules. The method we suggest is compared to two other popular prediction methods, a Naive Bayes classifier and a neural network, and appears to be superior to both, both in terms of learning quickly and being able to reach a high prediction accuracy.
- Home
- Academic Articles
- Mining Survey Data
S | M | T | W | T | F | S |
---|---|---|---|---|---|---|
1 | 2 | 3 | ||||
4 | 5 | 6 | 7 | 8 | 9 | 10 |
11 | 12 | 13 | 14 | 15 | 16 | 17 |
18 | 19 | 20 | 21 | 22 | 23 | 24 |
25 | 26 | 27 | 28 | 29 | 30 |
List of events
Recent Posts
- เชิญเป็นวิทยากร July 2, 2022
- วิทยากร [19-20 เมษายน 2565] April 6, 2022
- สาขาวิชาเทคโนโลยีและสื่อสารการศึกษา April 6, 2022
- วิทยากร [14 พฤศจิกายน 2564] November 4, 2021
- เกียรติบัตร: RDM The Series Season 2 August 27, 2021
- วิทยากร [24-27 สิงหาคม 2564] August 25, 2021
- Practical Data Science and Machine Learning#1 August 18, 2021