Data Week - 4

Suppose these values were categorized like in the Picture without pre-defined labels (unsupervised learning). What algorithm might be used?

• Linear Regression

• KMeans Clustering

• Decision Tree Classifier

Our classifier model could perform in accuracies of 0.95 and 0.80 in train and test sets, respectively.

Which statement is true?

• There is under-fitting

• There is over-fitting

• The model generalized well

Our classifier model could perform in accuracies of 0.45 and 0.42 in train and test sets, respectively.

Which statement is true?

• There is under-fitting

• There is over-fitting

What is this equation called?

• Precision

• Recall

What is this equation called?

• Precision

• Recall

Our classifier model could predict 80 out of 100 samples correctly in total.

What is the accuracy score?

• 0.2

• 0.8

Our classifier model could predict 80 out of 100 samples wrongly in total.

What is the accuracy score?

• 0.2

• 0.8

Our classifier model could predict 80 samples from class – 1 correctly

while 20 samples were wrongly evaluated as class – 1.

What are precision and recall scores for the respective class based on equations on the Picture?

• Precision is 0.2 and recall is 0.8

• Precision is 0.8 and recall is 0.2

• Recall is 0.8 but not enough information was given for precision.

• Precision is 0.8 but not enough information was given for recall.

Consider this as a binary classification.

Our classifier model could predict 80 samples from class – 1 correctly

while 20 samples were wrongly evaluated as class – 0.

What are precision and recall scores for the class – 1 based on equations on the Picture?

• Precision is 0.2 and recall is 0.8

• Precision is 0.8 and recall is 0.2

• Recall is 0.8 but not enough information was given for precision.

• Precision is 0.8 but not enough information was given for recall.

What is this equation called?

• Mean squared error

• Mean absolute error

• Root mean square error