On kNN Classification and Local Feature Based Similarity Functions - Agents and Artificial Intelligence - page 229

Fig. 1. Example images taken from the dataset

(104 photos); Camposanto Monumentale (exterior) (46 photos); Camposanto Monu-
mentale (field) (113 photos); Camposanto Monumentale (portico) (138 photos); Chiesa
della Spina (112 photos); Palazzo della Carovana (101 photos); Palazzo dell'Orologio
(92 photos); Guelph tower (71 photos); Basilica of San Piero (48 photos); Certosa (53
photos).
In order to build and evaluating a classifier for these classes, we divided the dataset
in a training set ( Tr ) consisting of 226 photos (approximately 20% of the dataset) and
a test set ( Te ) consisting of 921 (approximately 80% of the dataset). The image resolu-
tion used for feature extraction is the standard resolution used by Flickr i.e., maximum
between width and height equal to 500 pixels.
The total number of local features extracted by the SIFT and SURF detectors were
about 1,000,000 and 500,000 respectively.

6.2 Performance Measures

For evaluating the effectiveness of the classifiers in classifying the documents of the
test set we use the micro-averaged accuracy and micro- and macro-averaged precision ,
recall and F 1 .
Micro-averaged values are calculated by constructing a global contingency table and
then calculating the measures using these sums. In contrast macro-averaged scores are
calculated by first calculating each measure for each category and then taking the av-
erage of these. In most of the cases we reported the micro-averaged values for each
measure.
Precision is defined as the ratio between correctly predicted and the overall predicted
documents for a specific class. Recall is the ratio between correctly predicted and the
overall actual documents for a specific class. F 1 is the harmonic mean of precision and
recall .
Note that for the single-label classification task, micro-averaged accuracy is defined
as the number of documents correctly classified divided by the total number of docu-
ments in the test set and it is equivalent to the micro-averaged precision , recall and F 1
scores.