Table 14 Accuracy measures derived for the Solidago class for the OCSVM classification scenarios used in this study (strategy 1).

From: Harnessing remote sensing and machine learning techniques for detecting and monitoring the invasion of goldenrod invasive species

Scenario

Sentinel-2

PlanetScope

UA

PA

F1-score

OA

UA

PA

F1-score

OA

1

0.708

0.800

0.751

0.949

0.555

0.892

0.685

0.921

2

0.677

0.840

0.750

0.946

0.764

0.870

0.814

0.962

3

0.857

0.780

0.817

0.967

0.795

0.720

0.756

0.955

4

0.822

0.880

0.851

0.971

0.579

0.818

0.678

0.926

5

0.795

0.644

0.712

0.950

0.705

0.444

0.545

0.929

6

0.824

0.702

0.758

0.957

0.672

0.540

0.599

0.931

7

0.723

0.762

0.742

0.949

0.709

0.620

0.662

0.921

8

0.826

0.606

0.699

0.950

0.646

0.616

0.631

0.931

9

0.950

0.728

0.824

0.970

0.813

0.584

0.680

0.947

10

0.853

0.682

0.758

0.958

0.813

0.504

0.622

0.941

11

0.979

0.746

0.847

0.974

0.995

0.802

0.888

0.958

12

0.940

0.816

0.874

0.977

0.998

0.826

0.904

0.983

13

0.905

0.836

0.869

0.976

0.905

0.824

0.863

0.975

14

0.891

0.848

0.869

0.976

0.819

0.808

0.814

0.965

15

0.811

0.902

0.854

0.971

0.739

0.794

0.766

0.954

16

0.937

0.868

0.901

0.982

0.823

0.826

0.824

0.966

17

0.957

0.836

0.892

0.981

0.964

0.860

0.909

0.984

  1. UA: knowledge based; PA: Random Forest; OA = Overall accuracy.
Back to article page