1. Associative Classification (AC) Mining for A Personnel Scheduling Problem
Fadi Thabtah

2. Trainer scheduling problem
Schedule
Courses (events)
Resources
Locations
Staff (trainers)
Timeslots

3. Trainer scheduling problem
Assigning a number of training courses (events) to a limited number of training staff, locations, and timeslots
Each course has a numerical priority value
Each trainer is penalised depending on the travel distance

4. Objective Function MAX Total priority for scheduled events
Total penalty for training staff
MAX

5. Hyperheuristic approach
Operates at a higher level of abstraction than metaheuristics
You may think of it as a supervisor that manages the choice of simple local search neighbourhoods (low-level heuristics) at any time

6. Low-level heuristics Problem-oriented
Represent simple methods used by human experts
Easy to implement
Examples:
Add new event to the schedule
Swap two events in the schedule
Replace one event in the schedule by another

7. Hyperheuristic
Current solution
Low Level
Heuristic 1
Low Level
Heuristic 2
Low Level
Heuristic 3
Perturbed solution

8. Building a Schedule using A hyperheuristic
Initial solution
Hyperheuristic algorithm
Set of low-level heuristics
Selected low-level heuristic
Perturbed solution
CPU time
Objective value
Current solution (according to acceptance criterion)

9. Advantages of hyperheuristics
Cheap and fast to implement
Produce solutions of good quality (comparable to those obtained by hard-to-implement metaheuristic methods)
Require limited domain-specific knowledge
Robustness: can be effectively applied to a wide range of problems and problem instances

10. Current Hyperheuristics Approaches
Simple hyperheuristics (Cowling et al., )
Choice-function-based (Cowling et al., 2001 – 2002)
Based on genetic algorithms (Cowling et al., 2002; Han et al., 2002)
Hybrid Hyperheuristics.
(Cowling, Chakhlevitch )

11. Why Data Mining Scenario:
While constructing the solution of the scheduling problem, the hyperheuristic manages the choice of appropriate LLH in each choice point, therefore an expert decision maker is needed (Classification). Two approaches:
Learn the performance of LLH from past schedules to predict appropriate LLH in current one
While constructing schedule learn and predict
LLH Or what so called, Learn “On-the-fly”

12. Classification Algorithm
Classification : A Two-Step Process
1. Classifier building: Describing a set of predetermined classes
2. Classifier usage:
Calculate error rate
If Error rate is acceptable, then apply the classifier to test data
Classification Algorithm
Training Data
Class/
LLH
Test Data
RowIds A1 A2 1 x1 y1 c1 2 y2 c2 3 4 5 x2 6 7 y3 8 9 y4 10 x3
RowId A1 A2
Class 1 x1 y1 2 x2 y4 3
Classification Rules

13. Learning the Performance of LLH
(Hyperheuristic Solution)
Applied K
times
llh
oldpriority
newpriority
oldpenalty
newpenalty
applied 1
71954
72054
793
790 2 20
71054
761 27 37 43 47 58 68 74
Data Mining
Techniques
Produce
Derived
Hyperheuristic
Algorithm
Guide
Rules Set
(If/Then)

14. Association Rules Mining
Advantages:
Items shelving
Sales promotions
Future planning
Strong tool that aims to find relationships between variables in a database.
Its applied widely especially in market basket analysis in order to infer items from the presence of other items in the customer’s shopping cart
Example : if a customer buys milk, what is the probability that he/she buys cereal as well?
Unlike classification, the target class is not pre-specified in association rule mining.
Transactional Database
Transaction Id
Items
Time 12
bread, milk, juice
10:12 13
bread, juice, milk
12:13 14
milk, beer, bread, juice
13:22 15
bread, eggs, milk
13:26 16
beer, basket, bread, juice
15:11

15. Associative Classification (AC)
Special case of association rule that considers only the class label as a consequent of a rule.
Derive a set of class association rules from the training data set which satisfy certain user-constraints, i.e support and confidence thresholds.
To discover the correlations between objects and class labels.
Ex:
CBA
CPAR
CMAR

16. Attribute values that pass support threshold Class Association Rules
AC Steps
Training
Data
Associative classification Algorithm
Frequent Ruleitems:
Attribute values that pass support threshold
user
Class Association Rules

17. Rule support and confidence
Given a training data set T, for a rule
The support of R, denoted as sup(R) , is the number of objects in T matching R condition and having a class label c
The confidence of R , denoted as conf(R), is the the number of objects matching R condition and having class label c over the number of objects matching R condition
Any Item has a support larger than the user minimum support is called frequent itemset

18. Current Developed Techniques
MCAR (Thabtah et al., Pceeding of the 3rd IEEE International Conference on Computer Systems and Applications (pp. 1-7)
MMAC (Thabtah, et al., Journal of Knowledge and Information System (2006)00:1-21.
MCAR Characteristics:
Combinations of two general data mining approaches, i.e. (association rule, classification)
Suitable for traditional classification problems
Employs a new method of finding the rules
MMACC characteristics:
Produces classifiers of the form:
that are suitable to not only traditional binary classification problems but also useful to multi-class labels problems such as Medical Diagnoses and Text Classification.
Presents three Evaluation Accuracy measures

19. Data and Experiments Supp=5%, confidence=40%
Learning Approach : Learn the performance of LLH
from past schedules to predict appropriate LLH in
current one
Supp=5%, confidence=40%
Number of datasets : UCI data and 9 solutions
Of the training scheduling problem
Algorithms used:
CBA (AC algorithm)
MMAC (AC algorithm)
Decision Tree algorithms (C4.5)
Covering algorithms (RIPPER)
Hybrid Classification algorithm (PART)

20. Relative prediction accuracy in term of PART for the Accuracy Measures of MMAC algorithm

21. Relative prediction accuracy in term of CBA for the Accuracy Measures of MMAC algorithm

22. Number of Rules of CBA, PART and Top-label

23. Accuracy (%) for PART, RIPPER, CBA and MMAC on UCI data sets

24. Comparison between AC algorithms on 12 UCI data sets

25. MCAR vs. CBA and C4.5 On UCI data sets
Classification Accuracy %
Classifier Number of Rules
Dataset
MCAR
CBA
C4.5
Tic-tac
100
83.61 26 25 95
Balloon 3
Contact
93.33
83.33 9 6 4
Led7
72.32
71.1
73.34
192 50 37
Breast-cancer
71.61
69.66
75.52 71 45
Weather 5
Heart-c
80.4
79.87
78.12 72 44 12
Heart-s
81.35
79.2
81.29 31 22 2
Lymph
78.5
75.09
83.78 48 38
Mushroom
97.65
94.18 33
primary-tumour
40.5
25.47
42.47 28 1 23
Vote
90.1
86.91
88.27 84 40
CRX
83.05
85.31
80.72 97 43 54
Sick
93.88
93.9
93.87 17 10
Credit-Card
70.26
70.4
71.8
162
116

26. Conclusions
Associative classification is a promising approach in data mining
Since more than LLHs could improve the objective function in the hyperheuristic, we need a multi-label rules in the classifier
Associative classifiers produce more accurate classification models than traditional classification algorithms such as decision trees and rule induction approaches
One challenge in associative classification is the exponential growth of rules, therefore pruning becomes essential

27. Future Work
Constructing a hyperheuristic approach for the personnel scheduling problem
Investigating the use of multi-class labels classification algorithms with a hyperheuristic
Implementing of a new data mining techniques based on dynamic learning suitable for scheduling and optimization problem.
Investigate rule pruning in AC mining

28. Questions ?