| 2 | |
| 3 | Unless your dataset is exceptionally large (>400) you will have more dimensions in you dataset than samples. If you need to check, the number of dimensions is twice the number of points in your template (i.e. x and y coordinates in the picture), strictly speaking you can subtract up to 6 as a result of the alignment process but that only works if both you and the stats package understand rank-deficient matrices. In order to reduce that number of dimensions, and reduce the chance of over-fitting we use '''PCA''', fortunately shape PCA is built into Psychomorph in the form of an '''ASM''' (Active Shape Model). |
| 4 | |
| 5 | [wiki:recipes/PCA/explanation PCA (somewhat) explained] - A long winded incoherent rant on the topic. |
| 6 | |
| 7 | == The PCA Menu == |
| 8 | [wiki:recipes/PCA/BuildShapePCA Build Shape PCA (ASM)] – Build an Active Shape Model(ASM) from a set of templates[[BR]] |
| 9 | [wiki:recipes/PCA/BatchAnalyseShape Batch Analyse Shape] – Analyse a set of templates using an ASM and produce a set of PCA parameters[[BR]] |
| 10 | [wiki:recipes/PCA/BatchShapeReconstruct Batch Shape Reconstruct] – Reconstruct a set of templates from a set of PCA parameters using the ASM[[BR]] |
| 11 | ---[[BR]] |
| 12 | [wiki:recipes/PCA/BuildImagePCA Build Image PCI] – Build a Principal Components Image (PCI) model from a set of images[[BR]] |
| 13 | [wiki:recipes/PCA/BatchAnalyseImages Batch Analyse Images] – Analyse a set of images uinsg a PCI model and produce a set of PCA parameters[[BR]] |
| 14 | [wiki:recipes/PCA/BatchImageReconstruct Batch Image Reconstruct] – Reconstruct a set images from a PCI model and a set of PCA parameters.[[BR]] |
| 15 | PCA Options[[BR]] |
| 16 | -'''Mask Input Images''' – Mask images prior to bulding PCI model or analysing image sets.[[BR]] |
| 17 | -'''Mask Output Images''' – Mask images reconstructed from PCI model.[[BR]] |
| 18 | -'''Normalise Images''' – Pre-normalise images before building or analysing PCI model.[[BR]] |
| 19 | -'''Save Residual Images'''. – Save residuals from image analysis.[[BR]] |
| 20 | |
| 21 | |
| 22 | [[Image(average.jpg, width=40)]] |