Manifold learning can reduce a 100-dimensional dataset to 2D for visualization. 流形学习可以把100维的数据集降到二维用于可视化。
By assuming the data lie on a low-dimensional manifold, manifold learning helps uncover the underlying structure of complex observations like images or sensor readings. 通过假设数据位于低维流形上,流形学习有助于揭示图像或传感器读数等复杂观测背后的结构。