manifold.html

来自「一个由Mike Gashler完成的机器学习方面的includes neural」· HTML 代码 · 共 39 行

HTML
39
字号
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"><html><head></head><body><div style="text-align: center;"><big style="font-weight: bold;"><big>ManifoldSculpting</big></big><br></div><br>This is a demo of a manifold learning algorithm.<br><br><span style="font-weight: bold;">Swiss Roll Demo</span> - Starts with acollection of points in 3D space on a 2D Swiss Roll manifold, andreduces the data to two dimensions.<br><br><span style="font-weight: bold;">Semi-supervised Demo</span> - This isthe same as the Swiss Roll Demo, but when it's done, click on the 'X'and it will randomly pick 5% of the data to use to supervise the nextpass. Then it will do it again. Note that it's much faster the secondtime because it has some supervision.<br><br><span style="font-weight: bold;">S-Curve Demo</span> - Starts withpoints in 3D space on an 2D S-Curve manifold, and reduces the data totwo dimensions.<br><br><span style="font-weight: bold;">Entwined Spirals Demo</span> - Startswith points in 3D space on a 1D manifold of entwined spirals, andreduces the data to one dimension.<br><br><span style="font-weight: bold;">Face Demo</span> - Takes a collectionof images and reduces the data to one dimension. This essentiallyorders the images according to whatever change is happening across theimages. (The images are displayed on several wrapped lines since thereare too many to fit on a single horizontal line.)<br><br><br></body></html>

⌨️ 快捷键说明

复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?