International Business Machines Corporation

Frederick J Scheibl

Harry Roger Kolar

Vasken Torossian

Julio Ortega

Quan G Cung

Ben Peter Yuhas

Kevin Eric Norsworthy

Attributes of a data set to be employed in generating a predictive model are analyzed based on entropy, chi-square, or similar statistical measure. A target group of samples exhibiting one or more desired attributes is identified, then remaining attribute values for the target group are compared to corresponding attribute values for the whole sample population. A subset of all available attributes is then selected from those attributes which exhibit, when comparing attribute values of target group samples to attribute values for the whole sample population, the greatest relative difference or divergence. This subset is employed to generate the predictive model. Efficiency in generating the predictive model and the accuracy of the resulting predictive model is improved, since fewer attributes are employed and less computational resources are required.

Method for computing models based on attributes selected by entropy

A method and system for representing an arbitrary parametric surface having one or more trimming polylines applied thereto. A quadrilateral mesh coextensive with the parametric surface is generated. The quadrilateral mesh has a plurality of edges and vertices coinciding with the line segments and points of the trimming polylines. In order to generate the quadrilateral mesh, a two-dimensional array of U,V values is defined, wherein points in the array are adjusted to include the points of the trimming polylines. After all of the points needing adjustment are adjusted, the points in the array are evaluated, thereby creating geometric coordinate values for each point in the array.

System and method for generating a trimmed parametric surface for

A computer graphics display system and method are described for rendering a scene formed of at least one geometric primitive as a pixel image having shadows produced by at least one defined light source. Multiple passes are made through the primitive data structure for each light source capable of producing shadows in the scene to be rendered. In a first pass, the scene is rendered to a frame buffer in the usual way, but using only the ambient component of the light specification, and a first Z-buffer is updated with the viewpoint Z value. For each defined light source (i), two additional passes (PASS 2i & 2i+1) through the data structure are required. In the first of these, a transformation matrix is set up in such a way that the viewpoint is moved to the position of the light source. The scene is then rendered in the usual way except that the frame buffer is not updated, and a second Z-buffer (light source view Z-buffer) is used instead of the first Z-buffer. In the next pass, the shaded image and shadows are generated in parallel using the content of the first and second Z-buffers. When the frame buffer is updated, it is accomplished in a cumulative manner with each computed intensity value due to a specific light source being added to any value already stored there. In this way, intensities resultant from each of multiple light sources are accumulated on the image.

Growing community of inventors

Rhinebeck, NY, United States of America

Frederick J Scheibl