Data Mining With Weka Heart Disease Dataset: 1 Problem Description

Data Mining with Weka
Heart Disease Dataset
1 Problem Description
The dataset used in this exercise is the heart disease dataset available in
heart-c.arff obtained from the UCI repository1 . This dataset describes
risk factors for heart disease. The attribute num represents the (binary) class
attribute: class <50 means no disease; class >50_1 indicates increased level of
heart disease.
The main aim of this exercise is to predict heart disease from the other at-
tributes in the dataset. Obviously, this is a classification problem. The software
to be used is Weka 3.6 . However, feel free to try any ideas you may have to
tackle the problem with any other software.
The description of this exercise is stepwise. Therefore, I hope you can get a
better understanding of the various aspects and questions involved in the KDD
process.
1.1 Data Understanding

The first step in approaching the problem is to get acquainted with the data.
Answering the following questions will help you to better understand the data.
The data file heart-c.arff contains some information about the data stored
in it. You can open it with a text editor.
Load the data file in Weka.
1. For each attribute find the following information.

(a) The attribute type, e.g. nominal, ordinal, numeric.
(b) Percentage of missing values in the data.
(c) Max, min, mean, standard deviation.
(d) Are there any records that have a value for the attribute that no
other record has?
(e) Study the histogram at the lower right and informally describe how
the attribute seems to influence the risk for heart disease. What does
1A link to UCI data repository is available from the course web page.
1
it mean the pop-up messages that appear when dragging the mouse
over the graphic?
(f) Are there any outliers for the attribute under consideration?
i. Investigate the possibility of using the Weka filter InterquartileRange
to detect outliers2 .
2. Switch to the Visualize tab on the upper part of the screen to visualize
2D-scatter plots for each pair of attributes.
(a) Which attributes seem to be the most/least linked to heart disease?
Summarize in a table your findings concerning the predictive value
of each attribute.
(b) Does any pair of attributes seem to be correlated?
3. Investigate also possible multivariate associations of attributes with the
class attribute, i.e. study scatter plots of two attributes X and Y and try
to identify possible dense heart disease areas (if any).
(a) If you find dense heart disease areas in any scatter plot then quan-
tify the heart disease rate in these areas with respect to the entire
data set.
1.2 Data Preprocessing

The second step is to preprocess the data such that the transformed data is in
a more suitable form for the mining algorithms.
1. Attribute selection.
Investigate the possibility of using the Weka filter AttributeSelection
for selecting a subset of attributes with good predicting capability. Then,
describe briefly the filter(s) you used and compare the results you obtained
with the conclusions you obtained in the previous section. Save the dataset
with the selected attributes in the file heart-c1.arff.
2. Handling missing values.
Consider the following methods for handling missing values and investigate
each possibility within Weka. Note that, as rule of thumb, if an attribute
has more than 5% missing values then the records should not be deleted
and it is advisable to impute values where data is missing, using a suitable
method.
(a) Replace the missing values by the attribute mean, if the attribute
is numeric. Otherwise, replace missing values by attribute mode (if
the attribute is categorical). Save the dataset you obtained without
missing values in the file heart-c2.arff.
2 See also the information about box plots available from course web page.
2
(b) Investigate the possibility of using (linear) regression to estimate the
missing values for each attribute. Save the dataset you obtained
without missing values in the file heart-c3.arff.
3. Eliminating outliers.
Eliminate the outlier records and save the dataset you obtained without
outliers in the file heart-c34.arff.
1.3 Mining the Data

The third step is to use some classifier algorithms available in Weka to discover
hidden patterns in the data.
You should repeat the steps described below for each of the datasets you
created during preprocessing, besides using also the original dataset (if possible).
1. Start with OneR classifier.

(a) What can you conclude? Compare your conclusions with your pre-
vious conclusions obtained in section 1.1.
(b) Compare the accuracy of the classifier on the training set with the
accuracy estimation obtained through 10 fold-cross validation. How
do you explain the difference (if any)?
2. Use JRip classifier, i.e. the Weka version of the rule classifier RIPPER.
(a) Build a classifier with and without rule pruning. Which one is prefer-
able? Motivate your answer.
(b) Describe the patterns you obtained and compare with your previous
conclusions.
3. Use J48 classifier, i.e. the Weka version of the decision tree classifier C4.5.
(a) Investigate the use of different J48s parameters such as pruning and
minimum number of records in the leaves.
(b) Describe the patterns you obtained and compare with your previous
conclusions.
1.4 Clustering Tendency

Investigate whether there is a clustering tendency in the dataset. You may start
by clustering the data with SimpleKMeans algorithm, for some 2 k 10.
1. Do not use the class attribute, num, for clustering.

2. Find a suitable value for k, i.e. the number of clusters you are going to
build. Justify your choice of k.
3. Use class to cluster evaluation and make sure that standard deviations are
also computed for numerical attributes.
3
4. Study the numerical measures displayed by Weka for each cluster. What
can you conclude?
5. Use Visualize cluster assignments and try to discover a description

for each cluster.
6. Investigate the possibility of building a classifier for finding rules describing
the clusters. Compare the results with your previous conclusions.
7. Investigate the possibility of using the cluster information to build a clas-
sifier for num. Compare your results with what you obtained in section 1.3.
Do you get a better classifier?
1.5 Predicting Performance

In the previous step you have built several models. Finally, you need to compare
the different models and describe your final conclusions.
1. Weka outputs several performance measures. Choose some of the perfor-

mance measures and motivate your choice.
2. Summarize in a table the performance measures for each classifier and

each dataset.
3. What can you conclude?
1.6 Conclusions
Describe your final conclusions and indicate which risk factors for heart disease
have you found in the data.

Data Mining With Weka Heart Disease Dataset: 1 Problem Description

Uploaded by

Data Mining With Weka Heart Disease Dataset: 1 Problem Description

Uploaded by

Data Mining with Weka

Heart Disease Dataset

1.1 Data Understanding

1. For each attribute find the following information.

1.2 Data Preprocessing

1.3 Mining the Data

1. Start with OneR classifier.

1.4 Clustering Tendency

1. Do not use the class attribute, num, for clustering.

5. Use Visualize cluster assignments and try to discover a description

1.5 Predicting Performance

1. Weka outputs several performance measures. Choose some of the perfor-

2. Summarize in a table the performance measures for each classifier and

3. What can you conclude?

You might also like