-1

I am using 'sigmoid' activation function for multilabel classification and 'binary_crossentropy'

enter code here

img2 = img1/255
img3 = np.expand_dims(img2, axis=0)
ys_ = model.predict(img3)[0]
print('prediction',ys_)
Dhruvi Raval
  • 1
  • 1
  • 1
    What do you call "exponential value"? – Dr. Snoopy Sep 08 '22 at 06:44
  • [1.15890624e-14 2.39819930e-14 1.98484312e-10 4.41572962e-21 1.09111424e-24 5.96640589e-14 5.91789338e-08 2.54316693e-16 2.71238976e-10 6.16806430e-31 1.56134477e-18 3.96505717e-10 4.02764526e-05 5.53413528e-08 3.41465324e-07 9.53362052e-17 9.98271227e-01 1.55949122e-21 1.03139763e-14 1.71468542e-24 1.96253078e-14 1.07717027e-14 1.16775120e-13 1.73787512e-20 3.60638236e-19 1.47866518e-23 2.64950585e-22 5.16854031e-21 1.67879680e-16 1.91249041e-04 6.05797083e-14 4.14683542e-20 1.32537676e-14 8.29024256e-20 8.38788011e-25 3.34224509e-10] – Dhruvi Raval Sep 08 '22 at 06:45
  • I want like this 0.99999994, 0.9923841, 0.7607714, 0.6342545, 0.003293626 – Dhruvi Raval Sep 08 '22 at 06:46
  • That is called scientific notation, its just a matter of configuring your print functionality. – Dr. Snoopy Sep 08 '22 at 06:55
  • Can you give the detail explanation – Dhruvi Raval Sep 08 '22 at 06:57
  • You did not say how you were printing these values, there is no detail in your question. – Dr. Snoopy Sep 08 '22 at 07:13

1 Answers1

0

Use

np.set_printoptions(suppress=True)
s510
  • 2,271
  • 11
  • 18