about
History of commands that showed on seminar
    seminar_history.m
Example for training options
	example1.m
Predefined learning example:
	fine_tuning_example2
Training from scratch 
	train_example.m
the same, but images was rotated, shifted and mixed for better result
	train_example2.m
Confusion chart example:
	fine_tuning_example_validation.m
Demo image for demonstrating augmentation
	augmentation_single_image.m
*.mat files:
pretrained networks, but last layers is without weigths
	alexnet_4_layers_pretrained
		(class other is eleminated)
	alexnet_with_weights_modified_for_5_classes
totally untrained networks:
	alexnet_without_learnables
	alexnet_without_weigths
	alexnet_5_classes_totally_untrained
other mat files:
	training results 
you can open mat files with load command, then look what inside with whos
e.g.
clear
load fine_tuning_example2.mat
whos