Scribd
Upload
129 views

Tutorial 3: Statistics With MATLAB

This tutorial presents several statistics techniques using Matlab's Statistics toolbox, including descriptive statistics, linear models, cluster analysis, and principal component analysis. Topics covered include calculating correlation coefficients, standard deviation, variance, percentiles, and range of data sets, performing regression analysis, hierarchical clustering on data, and principal components analysis for dimensionality reduction. An example application analyzes texture data from CT images to classify tissues using correlation analysis, dimensionality reduction, and clustering.

Uploaded by

Wahyu Joe Pradityo
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
129 views

Tutorial 3: Statistics With MATLAB

This tutorial presents several statistics techniques using Matlab's Statistics toolbox, including descriptive statistics, linear models, cluster analysis, and principal component analysis. Topics covered include calculating correlation coefficients, standard deviation, variance, percentiles, and range of data sets, performing regression analysis, hierarchical clustering on data, and principal components analysis for dimensionality reduction. An example application analyzes texture data from CT images to classify tissues using correlation analysis, dimensionality reduction, and clustering.

Uploaded by

Wahyu Joe Pradityo
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Tutorial 3: Statistics with Matlab Page 1 of 4 02/20/2004

Tutorial 3♣ :
Statistics with MATLAB
Daniela Raicu
[email protected]
School of Computer Science, Telecommunications, and Information Systems
DePaul University, Chicago, IL 60604

The purpose of this tutorial si to present several statistics techniques using Matlab Statistics toolbox. For this
tutorial, we assume that you know the basics of Matlab (covered in Tutorial 1) and the basics of statistics. The
tutorial purpose is to teach you how to use the Matlab built-in functions to calculate the statistics for different data
sets in different applications; the tutorial is intended for users running a professional version of MATLAB 6.5,
Release 13.
Topics discussed in this tutorial include:
1. Descriptive statistics
2. Linear Models
3. Cluster analysis
4. Principal component analysis

1. Descriptive statistics
A. Correlation coefficient for two variables: “corrcoef.m”
>> a =

1 1 3 4

» b= [1 2 2 3]

b=

1 2 2 3

» corrcoef(a,b)

ans =

1.0000 0.8165
0.8165 1.0000

Question: What large the value of ‘ans’ should be in order to obtain very strongly correlated variables?
B. Standard deviation of a variable: “std.m”
>> » a=[3 5 7 8 9 11]

a=

3 5 7 8 9 11

♣
Event Sponsor: Visual Computing Area Curriculum Quality of Instruction Council (QIC) grant
Tutorial 3: Statistics with Matlab Page 2 of 4 02/20/2004

» std(a)

ans =

2.8577

C. Variance of a variable: “var.m”


>> » var(a)

ans =

8.1667

D. Percentiles for a data set: “prctile.m”


>> » prctile(a,50)

ans =

7.5000

» prctile(a,25)

ans =

» prctile(a,75)

ans =

» prctile(a,1)

ans =

» prctile(a,100)

ans =

11

E. Range for a data set: “range.m”


>> range(a)

ans =
8

2. Linear models
Regression analysis: “regress.m”
Tutorial 3: Statistics with Matlab Page 3 of 4 02/20/2004

>> b=[13 15 15 17 18 20]'

b=

13
15
15
17
18
20

» a=[3 5 7 8 9 11]'

a=

3
5
7
8
9
11

» coef=regress(a,b)

coef =

0.4516

3. Cluster analysis
Hierarchical clustering on a set of data: “clusterdata.m”

» X=[1 1 2 3;7 8 9 7;1 3 2 1;10 9 11 9;2 2 1 1;3 4 1 2]

X=

1 1 2 3
7 8 9 7
1 3 2 1
10 9 11 9
2 2 1 1
3 4 1 2

» T = clusterdata(X,0.8)

T=

1
2
1
2
1
1
Tutorial 3: Statistics with Matlab Page 4 of 4 02/20/2004

T = CLUSTERDATA(X,CUTOFF) is the same as

Y = pdist(X,'euclid');
Z = linkage(Y, 'single');
T = cluster(Z, CUTOFF);

4. Principal Component Analysis (for dimensionality reduction)


Principal components analysis from raw data: “princomp.m”
» [pc, score, latent, tsquare]=princomp(X)

pc =

0.4961 -0.3697 0.6397 -0.4561


0.4316 -0.6706 -0.4003 0.4515
0.6032 0.4179 -0.5296 -0.4254
0.4514 0.4889 0.3874 0.6381

score =

-4.7824 2.0736 0.3947 0.2489


7.2433 0.0423 -0.7265 0.2473
-4.8221 -0.2454 -1.1806 -0.1243
11.2723 0.0763 0.5080 -0.2441
-5.3608 -0.3624 0.3889 -0.6064
-3.5502 -1.5843 0.6155 0.4786

latent =

53.3994
1.4018
0.5731
0.1590

tsquare =

4.1571
2.2893
3.0077
3.2088
3.2088
4.1284

5. Application
Apply correlation analysis, dimensionality reduction, and clustering techniques for the data containing the texture
characterization for 193 CT images. The features/attributes consist of 11 te xture descriptors calculated from the raw
data and used to classify the tissues’ textures of five human body organs: heart, backbone, liver, spleen, and kidneys.
The data is contained in the file run_length_des.m”.

You might also like