来自预训练 tensorflow model 的混淆矩阵

Question

I trained the mobilenet v1 model of the tensorflow to detect my own classes, but I need to obtain the confusion matrix of the results of the model from the files that were generated, however I cannot find any post that is useful to me to create the matrix of a tensorflow model.

Answer 1

try this

import matplotlib.pyplot as plt
from matplotlib.pyplot import imshow
from sklearn.metrics import confusion_matrix, classification_report
y_pred=[]
preds= model.predict etc.....
for i, p in enumerate (preds):
    p=pred[i]
    predicted_class_index=np.argmax(p)  # this the the predicted column with highest probability assuming you class_mode='categorical'
    y_pred.append(predicted_class_index) # this is a list of predictions
# now you need to create a list of the corresponding true labels
# how you get this depends on how you supplied the data to model.fit
# but somewhere you should have a list of labels - these are y_true
# then do
cm = confusion_matrix(y_true, y_pred )

if you used the train_gen=ImageDataGenerator.flow_from_directory and set shuffle=False, then y_true=train_gen.labels