01 PCA
generation of points via linear transform
calculation of their covariance matrix 
decomposition of the matrix
inverse transform
(the covariance.py has to be performed gradually, try it with own values of the linear transform)
visualization of multidimensional data via PCA
(vizualize dataset of iris parameters from the Iris dataset by pca_visualization.py)
02 eigenimages
generation of faces 
(try to identify meaning of eigenvectors)
03 SVM
classifier of iris flowers from the Iris dataset
(identify samples that are not classified correctly by the linear SVM on the visualized dataset - see 01.
try the same with the radial base function)
04 Cascade regressor and gradient boosting
(try on suitable cropped image with face)

help: https://docs.opencv.org/4.3.0/
SVM - https://towardsdatascience.com/support-vector-machine-python-example-d67d9b63f1c8
Kazemi detector of the facial landmarks (cascade regressor) http://www.robotika.sk/seminar-archiv.php#cvicenie7

Homework: Modify program viewdataset and anotate (just a part of, e.g. 50 pcs.)
faces from the faces directory to two categories: men and women. Employ eigenvectors for finding
feature vectors (e.g. with length 50) from the dataset. Train SVM classifier of sex, i.e. evaluate image
whether it is man or woman. Evaluate its accuracy.
