Data Mining
Information Systems Department
Faculty of Computers and Artificial Intelligence
Cairo University
Assignment 3
Classification
Requirements:
Using the programming language you prefer, write a program to implements the following
problem.
Your program should include a graphical user interface.
Your program should implemented under the following specifications:
a. Inputs:
i. A file with a set of transactions (Excel, text, etc…). (Hint. The file attached).
ii. Divide the data set into 2 subsets, 1st one will be 75% of the data and call it
“Training Set”, 2nd set will be 25% of the data, and call it” Testing set”
iii. The size of each data set will be determined as an input from the user.
iv. The Class Label column will be the last column in the file chosen.
b. Outputs
i. The accuracy of the model.
ii. The class labels for the data records provided by the user.
c. General Requirements:
i. The program should enable the user to select the file needed to be analyzed,
ii. The interface should enable user to select the percentage of the data needed to be
read from the input file e.g. if the file contains 100 records, and the user needs to
read 70% of the file then the classification should be done on 70 records only.
iii. Apply the Bayesian classifier and decision tree classifier, to build two (classifier
models) from the first set “Training set”.
iv. Apply the Bayesian classifier and decision tree classifier you built on the second
set “Testing Set” to calculate the accuracy of the classifiers.
v. Compare the results of 2 classifiers Bayesian and Decision tree.
d. Instructions:
i. Assignment should be done individually; copies will be graded to -5.
ii. Total grade is 2 marks.
iii. No late submissions are allowed.
iv. Discussion will be held with Eng. Shrouk during the office hours.
v. Deadline: 6-5-2026.
1
Data Mining
Information Systems Department
Faculty of Computers and Artificial Intelligence
Cairo University
Problem Description:
Problem Description
You will be given a dataset related to census income data. The dataset contains records of
individuals along with demographic and employment-related attributes such as age, workclass,
education, marital status, and occupation. The goal is to predict whether an individual earns more
than $50K per year or not.
Hint: Use the most relevant attributes that may affect the prediction result (see dataset
description).
For this assignment, you are required to apply both the Naive Bayes classifier and the Decision
Tree classifier to determine whether an individual’s income exceeds $50K per year or not, based
on a selected set of five features.
The class label is:
Income → whether the individual earns more than $50K per year
(binary: >50K, <=50K)
Data Description
The dataset is derived from census records and includes demographic and employment-related
information about individuals. It is commonly used for classification tasks to predict income
levels based on personal and professional attributes.
Feature Values Range Description
Feature 1 Age (numeric) Age of the individual
Feature 2 Workclass (categorical: e.g., Private, Self-employed, Government, etc.) Type of employment
Feature 3 Marital Status (categorical: e.g., Married, Single, Divorced, etc.) Marital status
Feature 4 Education (categorical: e.g., Bachelors, HS-grad, Masters, etc.) Education level
Feature 5 Occupation (categorical: e.g., Tech-support, Sales, Exec-managerial, etc.) Type of job
Class Income (binary: >50K, <=50K) Income category