原书出处
Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)
The last line should make clear the intuition: we establish a contrast between data samples and current model samples, then update the model to decrease this contrast, bringing the model closer to the data. Once x(i) and are identically distributed, this gradient will be zero (in expectation); we have perfectly fit the data and no further updates should be taken.
原书定位详情
来源:a19c7933-4eda-4176-aebd-30fe7e1ad292
正文位置:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c032.xhtml
EPUB SHA-256:60a7eebc43f5d9a675d776fed3736fbdaa791e55877023f2f8cba016a9b384f8