Archivo: Una-May O'Reilly at Berkeley

Descripción: From Data to Knowledge - 508 - Una-May O'Reilly Slides: http://lyra.berkeley.edu/CDIConf/pdfs... Una-May O'Reilly: "Knit: Integrating Large Scale Partial Cognitive Analyses of Data". A video from the UC Berkeley Conference: From Data to Knowledge: Machine-Learning with Real-time and Streaming Applications (May 7-11, 2012). Abstract Una-May O'Reilly (MIT, CSAIL) Increasingly humans are becoming an integral part of knowledge discovery systems. Their ability to process information based on the context and their perception allows them to discern unique and novel patterns. However these systems are vulnerable to noisy and inconsistent behavior of humans.Additionally they are limited by human's cognitive capacity in processing data in real time. Our goal is to develop real-time knowledge discovery techniques which extend/accelerate/amplify the value of analyst's locally discerned patterns. Our approach knits together local similarity sets without explicit features. It also actively presents data to different analysts to effectively derive a consensus of global similarity.
Título: Una-May O'Reilly at Berkeley
Créditos: From Data to Knowledge - 508 - Una-May O'Reilly at 0:36, cropped, brightened
Autor(a): ckleinastro
Términos de Uso: Creative Commons Attribution 3.0
Licencia: CC BY 3.0
Enlace de Licencia: https://creativecommons.org/licenses/by/3.0
¿Se exige la atribución?: Sí