-
Notifications
You must be signed in to change notification settings - Fork 0
/
config.yaml
54 lines (45 loc) · 3.16 KB
/
config.yaml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
# Config file for wing-based automatic odonate identification
# Paths ------------------------------------------------------------------------
image_path : 'D:/Desktop/All scan images' # Path to training/testing images
squares_path : 'D:/Squares' #'D:/Desktop/All Py Squares' # Path to/for square (preprocessed) images
features_path : 'features.csv' # Path to table containing extracted features
metadata_path : 'metadata.csv' # Path to table with species names, etc. for each file
clf_setup_path : 'classifier_setup.py' # Path to classifier setup file
model_path : 'models_for_id' # Path to saved species classif models
mask_model_path : 'models_for_masking' # Path to saved masking models
images_for_masking_model_path : 'D:/TOWD_test_images_for_wing_masking' # Path to images for getting masking training data
masking_data_path : 'data_for_masking_model.csv' # Path to data for training masking model
# Log paths
log_preproc_path : 'log_image_preprocessing.csv'
log_extract_path : 'log_feature_extraction.csv'
# Linux paths ------------------------------------------------------------------
image_lpath : '/media/sf_All_scan_images' # Path to training/testing images
squares_lpath : '/media/sf_Squares' #'D:/Squares' # Path to/for square (preprocessed) images
# Log paths
log_preproc_lpath : '/media/sf_odomatic-wings/log_image_preprocessing.csv'
log_extract_lpath : '/media/sf_odomatic-wings/log_feature_extraction.csv'
# Class assignment -------------------------------------------------------------
# Classes can be created from individuals' genus, genus+species or
# genus+species+sex. Individuals from small classes can be reassigned to a
# separate class, 'rare_unknown', according to minClassSize, and 'rare_unknown'
# may be discarded with extraClass=False.
groupClassBy : ['Genus','Species']#,'Sex'] # how to group individuals
minClassSize : 30 # Min no. of indivs a class must have, otherwise it's reassigned to rare_unknown
extraClass : True # Include the rare_unknown class?
# Training-testing set parsing -------------------------------------------------
# Proportion of individuals per class to use for training vs testing
ttf : 0.9 # Proportion of indivs per class to use in training set
# Whether to limit class size to no more than the median original class size
# This helps when `rare_unknown` class is much larger than other classes
chop_large_classes : True
# Feature processing -----------------------------------------------------------
n_components : 200 # No. of principal components to keep during ordination
# Classification algorithm -----------------------------------------------------
# Which classification algorithm to use (refer to file in `clf_setup_path`)
clfMethod : 'rf' # Random Forest, calibrated & optimized
#clfMethod : 'nb' # Naive Bayes, calibrated, can't be optimized
#clfMethod : 'svm_rbf_g' # SVM with RBF kernel, optimized for genus
#clfMethod : 'svm_rbf_gs' # SVM with RBF kernel, optimized for genus+species or genus+species+sex
#clfMethod : 'svm_lin' # SVM with linear kernel, optimized
#clfMethod : 'nn_log' # Bernoulli RBM NeuralNetwork + Logistic Regression, calibrated & optimized
#clfMethod : 'nn_svm' # Bernoulli RBM NeuralNetwork + Linear SVM