Database Assignment
Due: November 13, 2009
- Give an example of a serializable but not recoverable schedule
- Give an example of a cascadeless but not serializable schedule
- Give an example of a scheudle that is neither serializable or recoverable
- Give an example of a schedule that is both casecadeless and serializable
- Show the results of applying the k-means algorithm on the first 20 prime numbers with 2 clusters.
- Using the items milk, bread and peanut butter, provide a set of transactions such that milk is associated with bread, peanut butter is associated with bread, but the set milk and peanut butter is NOT associated with bread. Use 50% as your support threshold. Defend your answer.
- Problem 18.12 on page 755
- Consider the table below. Predict User 1's rating of Items A and B based on user similarity (1 nearest neighbor).
- Consider the table below. Predict User 1's rating of Items A and B based on item similarity (2 nearest neighbors).
- Write a joke that will make me laugh. Compare and contrast classification against collaborative filtering on selecting that joke (100 word maximum).
| User 1 | User 2 | User 3 |
| Item A | | 1 | 1 |
| Item B | | 2 | 1 |
| Item C | 1 | 3 | 1 |
| Item D | 4 | 4 | 1 |
| Item E | 3 | 3 | 2 |