There are many problems associated with analyzing data sets that contain missing data. However, there are various techniques to 'fill in,' or impute, missing data values with reasonable estimates based on the characteristics of the data itself and on the patterns of 'missingness.' Generally, techniques appropriate for imputing missing values in multivariate normal data and not as useful when applied to non-multivariate-normal data. This Visualization and Imputation of Missing Data course focuses on understanding patterns of 'missingness' in a data sample, especially non-multivariate-normal data sets, and teaches one to use various appropriate imputation techniques to "fill in" the missing data. Using the VIM and VIMGUI packages in R, the course also teaches how to create dozens of different and unique visualizations to better understand existing patterns of both the missing and imputed data in your samples.
The course teaches both the concepts and provides software to apply the latest non-multivariate-normal-friendly data imputation techniques, including: (1) Hot-Deck imputation: the sequential and random hot-deck algorithm; (2) the distance-based, k-nearest neighbor imputation approach; (3) individual, regression-based imputation; and (4) the iterative, model-based, stepwise regression imputation technique with both standard and robust methods (the IRMI algorithm). Furthermore, the course trains one to recognize the patterns of missingness using many vibrant and varied visualizations of the missing data patterns created by the professional VIMGUI software included in the course materials and made available to all course participants.
This course is useful to anyone who regularly analyzes large or small data sets that may contain missing data. This includes graduate students and faculty engaged in empirical research and working professionals who are engaged in quantitative research and/or data analysis. The visualizations that are taught are especially useful to understand the types of data missingness that may be present in your data and consequently, how best to deal with this missing data using imputation. The course includes the means to apply the appropriate imputation techniques, especially for non-multivariate-normal sets of data which tend to be most problematic to impute.
The course author provides free-of-charge with the course materials his own unique VIMGUI toolbar developed in the RGtk2 visualization programming language in R. However, please note that both the R-provided VIMGUI package (developed in RGtk2), as well as the course author's provided VIMGUI toolbar application (also developed in RGtk2) may have some problems starting up properly on a Mac computer. So if you only have a Mac available to you, you may have some initial difficulties getting the applications to run properly.
Missing data are a common occurrence and can have a significant effect on the conclusions that can be drawn from the data. Missing datacan occur because of nonresponse: no information is provided for several items or no information is provided for a whole unit.
In statistics, imputation is the process of replacing missing data with substituted values. When substituting for a data point, it is known as "unit imputation"; when substituting for a component of a data point, it is known as "item imputation". Because missing data can create problems for analyzing data, imputation is seen as a way to avoid pitfalls involved with listwise deletion of cases that have missing values. That is to say, when one or more values are missing for a case, most statistical packages default to discarding any case that has a missing value, which may introduce bias or affect the representativeness of the results. Imputation preserves all cases by replacing missing data with an estimated value based on other available information. Once all missing values have been imputed, the data set can then be analysed using standard techniques for complete data
In pattern recognition, the k-Nearest Neighbors algorithm (or k-NN for short) is a non-parametric method used for classification and regression. In both cases, the input consists of the k closest training examples in thefeature space. The output depends on whether k-NN is used for classification or regression:
k-NN is a type of instance-based learning, or lazy learning, where the function is only approximated locally and all computation is deferred until classification. The k-NN algorithm is among the simplest of all machine learning algorithms.
Both for classification and regression, it can be useful to assign weight to the contributions of the neighbors, so that the nearer neighbors contribute more to the average than the more distant ones. For example, a common weighting scheme consists in giving each neighbor a weight of 1/d, where d is the distance to the neighbor.
The neighbors are taken from a set of objects for which the class (for k-NN classification) or the object property value (for k-NN regression) is known. This can be thought of as the training set for the algorithm, though no explicit training step is required.
A shortcoming of the k-NN algorithm is that it is sensitive to the local structure of the data.
Dr. Geoffrey Hubona held full-time tenure-track, and tenured, assistant and associate professor faculty positions at 3 major state universities in the Eastern United States from 1993-2010. In these positions, he taught dozens of various statistics, business information systems, and computer science courses to undergraduate, master's and Ph.D. students. He earned a Ph.D. in Business Administration (Information Systems and Computer Science) from the University of South Florida (USF) in Tampa, FL; an MA in Economics, also from USF; an MBA in Finance from George Mason University in Fairfax, VA; and a BA in Psychology from the University of Virginia in Charlottesville, VA. He is the founder of the Georgia R School (2010-2014) and of R-Courseware (2014-Present), online educational organizations that teach research methods and quantitative analysis techniques. These research methods techniques include linear and non-linear modeling, multivariate methods, data mining, programming and simulation, and structural equation modeling and partial least squares (PLS) path modeling.