Difference between revisions of "T-Distributed Stochastic Neighbor Embedding (t-SNE)"
m |
m |
||
| Line 19: | Line 19: | ||
* [http://projector.tensorflow.org/ Embedding Projector] | * [http://projector.tensorflow.org/ Embedding Projector] | ||
| − | + | a machine learning algorithm for visualization developed by Laurens van der Maaten and [[Creatives#Geoffrey Hinton |Geoffrey Hinton]]. It is a nonlinear dimensionality reduction technique well-suited for embedding high-dimensional data for visualization in a low-dimensional space of two or three dimensions. Specifically, it models each high-dimensional object by a two- or three-dimensional point in such a way that similar objects are modeled by nearby points and dissimilar objects are modeled by distant points with high probability. [http://en.wikipedia.org/wiki/T-distributed_stochastic_neighbor_embedding Wikipedia] | |
<youtube>wvsE8jm1GzE</youtube> | <youtube>wvsE8jm1GzE</youtube> | ||
Revision as of 18:13, 20 August 2020
Youtube search... ...Google search
- Principal Component Analysis (PCA)
- Embedding
- Dimensional Reduction Algorithms
- Softmax
- Pooling / Sub-sampling: Max, Mean
- (Deep) Convolutional Neural Network (DCNN/CNN)
- Seven Techniques for Dimensionality Reduction | KNIME
- Principal Component Analysis (PCA) ...linear
- Embedding Projector
a machine learning algorithm for visualization developed by Laurens van der Maaten and Geoffrey Hinton. It is a nonlinear dimensionality reduction technique well-suited for embedding high-dimensional data for visualization in a low-dimensional space of two or three dimensions. Specifically, it models each high-dimensional object by a two- or three-dimensional point in such a way that similar objects are modeled by nearby points and dissimilar objects are modeled by distant points with high probability. Wikipedia
