View article

Antonio Criminisi, Jamie Shotton, Ender Konukoglu

2011/10/28

Microsoft Research Cambridge, Tech. Rep. MSRTR-2011-114

5

6

12

Diapositiva 1 Page 1 Decision Forests for Classication, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning A. Criminisi1, J. Shotton2 and E. Konukoglu3 1 Microsoft Research Ltd, 7 JJ Thomson Ave, Cambridge, CB3 0FB, UK 2 Microsoft Research Ltd, 7 JJ Thomson Ave, Cambridge, CB3 0FB, UK 3 Microsoft Research Ltd, 7 JJ Thomson Ave, Cambridge, CB3 0FB, UK Josu Maiora Page 2 1. Overview and scope ❑ Unified, efficient model of random decision forests ▪ Applications • Scene recognition from photographs, • Object recognition in images, • Automatic diagnosis from radiological scans • Semantic text parsing. ❑ A brief literature survey ▪ Breinman ▪ “C4.5” of Quinlan ▪ In this early work trees are used as individual entities. However, recently it has emerged how using an ensemble of learners (eg weak classiffiers) yields greater accuracy and generalization.Grupo de Inteligencia …