论文部分内容阅读
Current interaction with ma-chine learning systems are mostly batch-based.A system takes a set of anno-tated data as input,and re-turns a trained model.This model is problematic be-cause the annotation process is labor intensive and the training process works as a black box,making it difficult to understand and control.We are developing techniques to address these problems by introducing rich visual inter-action techniques to help users understand and control machine learning systems.In this talk,I will present some of the results from our group in computer graphics.