Uploads by Njmurdoc
Jump to navigation
Jump to search
This special page shows all uploaded files.
| Date | Name | Thumbnail | Size | Description | Versions |
|---|---|---|---|---|---|
| 16:40, 26 November 2009 | Sep2.png (file) |  |
37 KB | A simple transformation to Euclidean distance from the origin, and it is possible to put a hyperplane through the data that separates it perfectly into two classes. | 1 |
| 16:35, 26 November 2009 | Unsep.png (file) |  |
38 KB | A set of points in 2d which no hyperplane can separate with good accuracy. | 1 |
| 16:20, 26 November 2009 | Point 2d.png (file) |  |
13 KB | In 2 dimensions, a point is trapped by a rectangle (shown in red). | 1 |
| 16:19, 26 November 2009 | Point 3d.png (file) |  |
21 KB | A point moving in 3d space can easily escape a 2d prison (shown in red). | 1 |
| 16:24, 8 November 2009 | Curvefitting-rex2.png (file) |  |
11 KB | Fitting degree 1, 2, and 10 polynomials, and trigonometric functions to a randomly generated quadratic sample. Black points represent training data, teal represent test data. Y values follow the same function, but x values of training and test data have | 1 |
| 16:01, 8 November 2009 | Curvefitting-rex.png (file) |  |
10 KB | Fitting degree 1, 2, and 10 polynomials to a randomly generated sample. Black points represent training data, teal represent test data. Y values follow the same function, but x values of training and test data have means 0 and 1 to help separate them. | 1 |
| 14:45, 9 October 2009 | Pca-fda1 low.png (file) |  |
58 KB | Multivariate normal distributions centred at (1,1) and (5,3), showing PCA and FDA primary dimensions. | 1 |
