Code for finding out accuracy of test set

#1

Can you please provide the code for finding out the accuracy of test set in the MNIST handwritten project using KNN

#2

tejas simply count the number of cases in which , test case value is equal to the generated output using simple loop.Then divide it by total number of test cases.
accuracy = (total number of cases in which test value equals generated value)/(total number of test cases)

#3

This is the code that I wrote ,

correctCase = 0
for i in X_test:
pred = knn(X_train,Y_train,X_test[i])
if pred == Y_test[i]:
correctCase += 1

accuracy = (correctCase) / (len(X_test))
print(accuracy)

But its giving a value error that says “the truth value of an array with more than one element is ambiguous. use a.any() or a.all() ​”

Please help

#4

hi tejas;
I appreciate ur efforts.
Plz use loop in different way to get results
for i in range(X_test.shape[0]):
if pred == Y_test[i]
because for i in X_test is iterating over X_test iterating as an object and not a number that we are using to iterate over index of Y_test.
that’s why erorr occuring

1 Like