Association Rule Mining

1. Association Rule Mining: (15 points)

 

Given a transaction database for mining association rule as follows:

 

Database D

 

TID	Items
100	A C D
200	B C E
300	A B C E
400	B E

 

Please use Apriori algorithm to mine association rules with minimum support count = 2.

(Please show the derivation process step by step with candidate itemsets.)

 

1. Generating Classification Rules from a Decision Tree (50 points)

 

Given a database table containing weather data as follows:

 

Outlook	Temperature	Humidity	Windy	Class: Play
Sunny	Hot	High	False	No
Sunny	Hot	High	True	No
Overcast	Hot	High	False	Yes
Rainy	Mild	High	False	Yes
Rainy	Cool	Normal	False	Yes
Rainy	Cool	Normal	True	No
Overcast	Cool	Normal	True	Yes
Sunny	Mild	High	False	No
Sunny	Cool	Normal	False	Yes
Rainy	Mild	Normal	False	Yes
Sunny	Mild	Normal	True	Yes
Overcast	Mild	High	True	Yes
Overcast	Hot	Normal	False	Yes
Rainy	Mild	High	True	No

 

1. Please use the basic algorithm (version of ID3) for inducing a decision tree from the given training samples in the weather database table.

 

 

2. Please extract the classification rules from the generated decision tree in 1.

 

1. For the weather database table given in B, please predict a class label for the weather data by using naïve Bayesian classification approach (20 points).

 

The unknown samples to be classified are:

 

(Outlook = ‘Sunny’,  Temperature = ‘Mild’ , Humidity = ‘High’ ,  Windy = ‘False’)

 

(Outlook = ‘Sunny’,  Temperature = ‘Hot’ , Humidity = ‘High’ ,  Windy = ‘False’)

 

 

1. Classification and Characteristic Rule Derivation (50 points)

 

For the given conditions as follows:

 

1.)   The relation table:

 

Student

Name	Sex	Age	Birth_place	Major	Position	Salary
Carrie	female	33	Florida	DCTE	Instructor	$31,000
Stilwell	male	58	Michigan	DISS	Manager	$65,000
Nana	female	35	Japan	DCTE	Instructor	$35,000
O’Hare	male	35	Canada	DCS	Assistant Prof.	$45,000
Peabody	male	50	New York	DISS	CIO	$70,000
Juliana	female	68	California	DISS	CEO	$90,000
O’Neil	male	42	France	DCS	Lecturer	$50,000
Diana	female	51	Oregon	DCTE	Assist. Prof.	$49,000
Anderson	male	45	India	DCS	Instructor	$49,500
Christopher	male	42	New Mexico	DISS	Manager	$55,000
Cook	female	40	Illinois	DISS	System Analyst	$45,000
Kim Ming	female	38	South Korea	DCTE	Lecturer	$35,500
Donovan	male	32	Netherlands	DCS	Assist. Prof.	$48,000
George	male	42	South Korea	DCS	Instructor	$35,000
Donna	female	35	Texas	DCS	Programmer	$57,000
Mike	male	60	Ohio	DISS	Manager	$67,500
Lynn	female	55	Georgia	DISS	Manager	$60,000
Lisa	female	37	Italy	DCS	Programmer	$47,500
Sherry	female	46	Germany	DCS	Lecturer	$38,000
Robert	male	51	Kansas	DISS	CIO	$72,500

 

 

2.)  The concept hierarchy table:

 

Age:

 

{21 – 30 } Ì Young

{31 – 50 } Ì Mid-Age

{51 – 70 } Ì Old

 

Birth_Place:

 

{Canada, France, Germany , India, Italy, Japan, Netherlands, South Korea }Ì Foreign

{California, Florida, Georgia, Illinois, Kansas, Michigan, New York, Ohio, Oregon, Utah, Texas, New Mexico } Ì USA

 

Major:

 

{ DCS,  DISS,  DCTE }

 

Position:

 

{Instructor, Lecturer, Assistant Professor, Associate Professor, Professor} Ì Faculty

{CEO, CIO, Manager, Programmer, System Analyst} Ì Non-Faculty

 

Salary:

 

{ $20,000 – $30,000 }Ì Low

{ $30,001 – $50,000 }Ì Medium

{ $50,001 – $100,000 }Ì High

 

Please do the followings:

 

1. Please derive the quantitative classification and characteristic rules from the given relation table and the concept hierarchy table. The target class for the quantitative rule derivation is male.
2. Please indicate the sufficient and necessary conditions for these derived rules.
3. Please indicate which attribute should be removed during the derivation process.
4. Please indicate the threshold value T for the number of distinct values of each remaining attribute.
5. Please give the detailed intermediate tables in the derivation process, mark the tuples which overlap between the target class and contrast class, and make assumptions whenever necessary.
6. Please make reference to the following materials for the examination:

 

Jiawei Han, Yandong Cai, and Nick Cercone, “Data-Driven Discovery of Quantitative Rules in Relational Databases,” IEEE Transactions on Knowledge and Data Engineering, Vol. 5, No. 1, 1993, pp. 29-40.

 

Yandong Cai, Nick Cercone, and Jiawei Han, “An Attribute-Oriented Approach for Learning Classification Rules from Relational Databases,” in Proceedings of Sixth International Conference on Data Engineering, February 1990, pp. 281-288.

 

 

 

 

 

 

 

 

 

 

 

 

 

1. Please answer the following questions: (35 points)

 

(a)   What is the confidence for the rules ∅® A and A ® ∅? (10 points)

 

 

(b) Let c₁, c₂, and c₃ be the confidence values of the rules {p}®{q}, {p}®{q, r}, and {p, r}®{q}, respectively. If we assume that c₁, c₂, and c₃ have different values, what are the possible relationships that may exist among c₁, c₂, and c₃? Which rule has the lowest confidence?

(15 points)

 

 

(c)   Repeat the analysis in part (b) assuming that the rules have identical support. Which rule has the highest confidence? (10 points)

 

 

1. Please give the answers to the following questions:

 

1. What is the lower bound and upper bound in terms of the number of candidate sets generated by the Apriori association rule mining algorithm? (15 points)

 

2. What is the total number of possible association rules, which can be generated from a given data set that contains n items? Why? (25 points)

 

 

 

Is this the question you were looking for? If so, place your order here to get started!