Weka reads in a toy dataset called "segment challenge art" which you can find in your Weka installation folder or here. The task is to classify regions in outdoor Images into one of the seven classes: brickface, sky, foliage, cement, window, path and grass. Logistic regression (functions.Logistic class) is run using its default parameter settings and 10-fold cross validation. The performance output is shown below: www Detailed Accuracy By Class TP Rate FP Rate Precision Recall -Measure MCC ROC Area PAC Area Class 0.980 0.005 0.971 0.980 0.976 0.972 0.997 8.974 brickface 1.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 sky 0.923 0.014 0.914 0.923 0.919 0.906 0.994 0.953 foliage 0.918 0.00 0.967 0.918 0.942 0.932 0.985 0.962 cement 0.897 0.022 0.867 0.897 0.882 0.863 .987 8.921 window 1.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 path 1.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 grass Weighted Avg. 0.961 0.006 0.961 0.961 0.961 0.955 0.995 0.974 Confusion Matrix 01 a b c d e 1 0 201 02 2 0 220 0 1 @ 192 1 15 001 4 0 3 202 11 @ @ 2 015 4 183 0 0 @ @ @ 0 0 236 0 0 0 @ 2071 classified as a brickface bsky C foliage dcement e window f path g grass 01 Which class is worst recognized, i.e. with lowest recall? The "recall" is a performance metric defined as the number of correctly labelled samples among all samples that should have that label. brickface window foliage path