Unsupervised methods

Two examples of unsupervised classical pattern recognition methods are hierarchical cluster analysis (HCA) and principal components analysis (PCA). Unsupervised methods attempt to discover natural clusters within data sets. Both HCA and PCA cluster data. 

An alternative clustering method for variables is to use the correlation coefficient matrix in which each variable is considered as an object, characterized by the correlation coefficients to all other variables. PCA and other unsupervised methods can be applied to this matrix to obtain an insight into the similarities between the original variables. 

In contras to unsupervised methods, supervised pattern-recognition methods (Section 4.3) use class membership information in the calculations. The goal of these methods is to construct models using analytical measurements to predict class membership of future samples. Class location and sometimes shape are used in the calibration step to construct the models. In prediction, these moddsare applied to the analytical measurements of unknowu samples to predict dsss membership. 

The final two habits are not applied when using unsupervised methods. 

Like the unsupervised methods, the supervised methods discussed in this book are based on the assumption that samples that are chemically or physically similar wM be near each other in measurement (row) space. 

The two unsupervised methods examined are HCA and PCA. HCA calculates the interpoint distances between all of the rows and represents that information in the form of a two-dimensional plot called a dendrogram. PCA calculates a new axis system that maximally describes the variation in the data set. Our recommendation is to use both of the methods whe " they are available. HCA gives a broader view of the data and PCA can be used to further investigate samples and dusters that are highlighted in HCA. 

Pattern recognition can be classified according to several parameters. Below we discuss only the supervised/unsupervised dichotomy because it represents two different ways of analyzing hyperspectral data cubes. Unsupervised methods (cluster analysis) classify image pixels without calibration and with spectra only, in contrast to supervised classifications. Feature extraction methods [21] such as PCA or wavelet compression are often applied before cluster analysis. 

Contingency table analysis (Fig. 10.10, left), which can be considered as an unsupervised method, revealed, for example, the expected high correlation between Cu and Zn, and also an unexpected correlation between Mo and Sm. The same element combinations were highlighted by the induction tree results (Fig. 10.10,... 

It is, at this point, important to understand the difference between unsupervised methods and supervised methods. With the former, there is no indication given to the model creation program (e.g. PCA, self-organising maps) of where any of... 

Data mining methods can be generally divided into two types, unsupervised and supervised. Whereas unsupervised methods seek informative patterns, which directly display the interesting relationship among the data, supervised methods discoverpredictivepatterns, which can be used later to predict one or more attributes from the rest. 

Pattern Recognition. The application of computers to build descriptive or predictive models (i.e., find patterns) of information from input datasets. The techniques of pattern recognition overlap those used in statistics, chemometrics, and data mining, and include data display, description, and reduction, unsupervised methods such as cluster analy-... 

Selected Data Mining Techniques for Gene Expression 9.15.3.1 Unsupervised Methods in Bioinformatics... 

Principle components analysis (PCA), a form of factor analysis (FA), is one of the most common unsupervised methods used in the analysis of NMR data. Also known as Eigenanalysis or principal factor analysis (PEA), this method involves the transformation of data matrix D into an orthogonal basis set which describes the variance within the data set. The data matrix D can be described as the product of a scores matrix T, and a loading matrix P,... 

In complex systems where the number of groups to be separated during classification becomes larger, the performance of simple unsupervised methods (Section 3) degrades, requiring the use of more sophisticated supervised chemometric techniques. Additionally, in fields such a process NMR where there is a need for quantifying a component, the use of supervised methods becomes necessary. The different supervised methods described in the sections below have all been utilized in the chemometric analysis of NMR data for classification and/or quantitation. Examples utilizing these different techniques are discussed in Section 5. 

QSAR methods can be classihed in several ways. One approach is to look at the nature of the method, supervised versus unsupervised, where supervised methods use the activity values to create a predictive model from the descriptors and unsupervised methods model molecular similarity from descriptors, but do not use the activity values in the derivation of the model. Another way is to look at the nature of the relationship between activity and descriptors categorical versus continuous, or linear versus non-linear (Figure 23.1)... 

Although supervised and unsupervised methods have been in the past well-separated families, a number of hybrid approaches have appeared that combine both unsupervised and supervised classifiers for image segmentation purposes [61]. 

The earliest efforts to use spectroscopic methods for the diagnosis of disease used mostly a visual inspechon of the spectra and simple band intensity ratios to correlate spectral features and histopathology. In contrast, the results presented here uhhze supervised and unsupervised methods of mulhvariate statistics to maximize the spectral informahon used in the diagnoshc process. 

Exploratory methods are used not to test hypotheses but rather to get an overview of data. Various clustering methods and ordination are excellent tools for exploratory analysis of microarray data. These unsupervised methods do not require external class or group information. Clusters are generated purely based on the intrinsic similarity of the gene or sample expression profiles. No null hypothesis can be rejected, and p values are not generated to test statistical significance. Methods that... 

---