Cloud

Donated on 8/2/1989

Little Documentation

Dataset Characteristics

Multivariate, Image

Subject Area

Climate and Environment

Associated Tasks

Other

Feature Type

Real

# Instances

1024

# Features

10

Dataset Information

Additional Information

The data sets we propose to analyse are constituted of 1024 vectors, each vector includes 10 parameters. You can think of it as a 1024*10 matrix. To produce these vectors, we proceed as follows: 1. we start with two 512*512 AVHRR images (1 in the visible, 1 in the IR) 2. each images is divided in super-pixels 16*16 and in each super-pixel we compute a set of parameters: (a) visible: mean, max, min, mean distribution, contrast, entropy, second angular momentum (b) IR: mean, max, min The set of 10 parameters we picked to form the vectors is a compromised between various constraints. Actually we are still working on the choice of parameters for the data vectors. The data set I send you has not been normalized. The normalization of the data set is required by our classification scheme but that may not be true for yours. To normalize the data we compute the mean and standard deviation for each parameter on the entire data set then for each parameter of each vector we compute: Norm. value = (un-norm value - mean)/SD where mean = mean value for this particular parameter over the data set SD = standard deviation .....

Has Missing Values?

No

Variables Table

Variable NameRoleTypeDemographicDescriptionUnitsMissing Values
visible-meanFeatureIntegerno
visible-maxFeatureIntegerno
visible-minFeatureContinuousno
mean-distributionFeatureContinuousno
contrastFeatureContinuousno
entropyFeatureContinuousno
second-angular-momentumFeatureContinuousno
ir-meanFeatureIntegerno
ir-maxFeatureIntegerno
ir-minFeatureContinuousno

0 to 10 of 10

Reviews

There are no reviews for this dataset yet.

Login to Write a Review
Download
0 citations
21727 views

Creators

Philippe Collard

License

By using the UCI Machine Learning Repository, you acknowledge and accept the cookies and privacy practices used by the UCI Machine Learning Repository.

Read Policy