In this paper, we apply Constrained Maximum a Posteriori Linear Regression (CMAPLR) transformation on Universal Background Model (UBM) when characterizing each speaker with a supervector. We incorporate the covariance transformation parameters into the supervector in addition to the mean transformation parameters. Maximum Likelihood Linear Regression (MLLR) covariance transformation is adopted....

Empirical Similitude Method (ESM) is a technique developed to address the concerns of non – linear similitude. The process disconnects the scaling of material and shape response by assuming them to be independent to form a system transformation. Such a transformation requires modeling using three representative models instead of just one as is used in the Buckingham π theorem. In an effort to g...

We present abstraction techniques that transform a given non-linear dynamical system into a linear system or an algebraic system described by polynomials of bounded degree, such that, invariant properties of the resulting abstraction can be used to infer invariants for the original system. The abstraction techniques rely on a change-of-basis transformation that associates each state variable of...

A series of motion compensation algorithms is run on the challenge data including methods that optimize only a linear transformation, or a non-linear transformation, or both – first a linear and then a non-linear transformation. Methods that optimize a linear transformation run an initial segmentation of the area of interest around the left myocardium by means of an independent component analys...

Many methods for deriving vectorcardiographic (VCG) leads from 12-lead electrocardiogram (ECG) were published. All of them are based on a linear transformation. Differences are only in coefficients of transformation matrixes which are designed based on a model or on linear regression methods. This paper presents and assesses a new nonlinear transformation method based on artificial neural netwo...

When it becomes necessary to reduce the complexity of a classifier, dimensionality reduction can be an effective way to address classifier complexity. Linear Discriminant Analysis (LDA) is one approach to dimensionality reduction that makes use of a linear transformation matrix. The widely used Fisher’s LDA is “sub-optimal” when the sample class covariance matrices are unequal, meaning that ano...

We define the domain of a linear fractional transformation in a space of operators and show that both the affine automorphisms and the compositions of symmetries act transitively on these domains. Further, we show that Liouville’s theorem holds for domains of linear fractional transformations and, with an additional trace class condition, so does the Riemann removable singularities theorem. We ...

REPORTRAPPORT A linear-time transformation of linear inequalities into conjunctive normal form Abstract We present a technique that transforms any binary programming problem with integral coeecients to a satissability problem of propositional logic in linear time. Preliminary computational experience using this transformation, shows that a pure logical solver can be a valuable tool for solving ...