This course introduces principles and analyses related to data with categorical outcomes. This course will consider topics such as probability distributions with categorical data, contingency table analysis, the generalized linear model, logit models and loglinear models. Students are expected to: a) learn to select methods appropriate for a question of interest for data with a categorical outc...

Ling Guo. Randomization Based Privacy Preserving Categorical Data Analysis. Under the direction of Dr. Xintao Wu The success of data mining relies on the availability of high quality data. To ensure quality data mining, effective information sharing between organizations becomes a vital requirement in today’s society. Since data mining often involves sensitive information of individuals, the pu...

The data stream model has been defined for new classes of applications involving massive data being generated at a fast pace. Web click stream analysis and detection of network intrusions are two examples. Cluster analysis on data streams becomes more difficult, because the data objects in a data stream must be accessed in order and can be read only once or few times with limited resources. Rec...

Graphical methods for quantitative data are well-developed, and widely used in both data analysis (e.g., detecting outliers, verifying model assumptions) and data presentation. Graphical methods for categorical data, however, are only now being developed, and are not widely used. This paper outlines a general framework for data visualization methods in terms of communication goal (analysis vs. ...

Recent studies in classification have proposed ways of exploiting the association rule mining paradigm. These studies have performed extensive experiments to show their techniques to be both efficient and accurate. However, existing studies in this paradigm either do not provide any theoretical justification behind their approaches or assume independence between some parameters. In this thesis,...

We study the problem of ignorability in likelihood-based inference from incomplete categorical data. Two versions of the coarsened at random assumption (car) are distinguished, their compatibility with the parameter distinctness assumption is investigated, and several conditions for ignorability that do not require an extra parameter distinctness assumption are established. It is shown that car...

Dynamical systems approach for clustering categorical data have been studied by some authors [1]. However, the proposed dynamic algorithm cannot guarantee convergence, so that the execution may get into an in nite loop even for very simple data. We de ne a new conguration updating algorithm for clustering categorical data sets. Let us consider a relational table with k elds, each of which can a...

• Section 10.1 Proportions — Goodness-of-fit • Section 10.2 2 × 2 tables — test of equality of population proportions • Section 10.3 2 × 2 tables — test of independence of categorical variables • Section 10.4 2 × 2 tables — Fisher’s exact test • Section 10.5 r × k tables • Section 10.6 Applicability — when are the methods valid? • Section 10.7 Confidence intervals for differences in porportions...