Reference no: EM133055262
Project for Convolutional Neural Network (CNN)
Use any one of the image database attached below
Requirements:
1) Use 70% data for training and 30% data for testing/validation
2) Calculate Accuracy (ACC) defined as
ACC=(TP+TN)/(P+N)
Where TP=True Positive, TN=True Negative, P=Positive, N=Negative
3) Calculate Sensitivity (Sens) defined as
Sens=TP/P
4) Calculate Specificity (Spec) defined as
Spec=TN/N
5) Plot the results for Accuracy and Loss function
6) Calculate the Confusion Matrix. As an example for 4 classes,it looks like (TC=True Class, PC=Predicted Class)
|
|
PC1
|
PC2
|
PC3
|
PC4
|
|
TC1
|
%
|
%
|
%
|
%
|
|
TC2
|
%
|
%
|
%
|
%
|
|
TC3
|
%
|
%
|
%
|
%
|
|
TC4
|
%
|
%
|
%
|
%
|
Attachment:- Convolutional Neural Network.rar