ETOIK7VEDSRQX6JISLV7WWAW2O7UTBWV6MHTRKVTHARXOD77MX2QC
QPBH7QWC56HIYUJCCHKDNBN2D346UN3IOXDCIKSAATLIATLSW3YQC
NV7FXZ5QETWHE7EQHET5ZZUKH4UIAIRGQ42MR2IT5JCZDPRNEZRQC
3FAESP6NBZJHKN6VSGENUMZXPID6O5CGNQ3NHJC5Z3YXTAXJVWVAC
import MLBase
train("K1-9_original_set", 20, glob_pattern, true, 0.90, 64)
train("K1-9_original_set", 2, glob_pattern, true, 0.90, 64)
images = glob(glob_pattern) #|> shuffle! |> x -> x[1:1000]
images = glob(glob_pattern) |> shuffle! |> x -> x[1:10000]
label_to_index,
label_to_index
function evaluate(m, d, c=classes)
function evaluate(m, d, c)
pred = [] actual = []
pred = []
actual = []
pred = Int64[] actual = Int64[]
pred = Int64[]
actual = Int64[]
r=roc(actual, pred)
r=MLBase.roc(actual, pred)
@time eval, vcm, vroc = evaluate(model, test)
@time eval, vcm, vroc = evaluate(model, test, classes)
@time metric_train, train_confusion_matrix, train_roc = evaluate(model, train_sample)
@time metric_train, train_confusion_matrix, train_roc = evaluate(model, train_sample, classes)
@time metric_test, test_confusion_matrix, test_roc = evaluate(model, test)
@time metric_test, test_confusion_matrix, test_roc = evaluate(model, test, classes)