Season 2 #Spaic ML Azure Weekly Quiz

What is Classification ?

• One of the common task in ML

• When the expected predicted output is in the form of discrete values

• When the expected predicted output is in the form of continues values

• One of the common approaches in ML

The type of Classification are :

• Binary classification

• Regression to arbitrary value

• Multi-class, Multi-label classification

• Multi-class Single-label classification

Confusion matrix is a matrix where the predicted vs the actual value are placed to see how well the model performed

• True

• False

What are the evaluation metrics for Classification?

• Accuracy (basically the proportion of correct prediction)

• Precision ( the proportion of positive cases that are correctly predicted)

• Recall ( the proportion of actual positive cases (T) that are correctly predicted )

• R squared (How close the regression line to the true value)

There are many different Classification algorithms such as :

• Logistic regression

• SVM support vector machine

• Linear Regression

• Decision Forest Regression

To create a model we set the compute target :

• Before creating the model

• After the creation of the model

Split data component is used to split the data into 2 majors sets ( the training and the testing sets)

• True

• False

What do we set in the training model component (edit column)?

• The input column used for the training

• The target column to predict

• All the above choice

We use the score model component :

• To evaluate the performance of the model

• To predict the result of the testing dataset

• All the above choice

What are the evaluation model chart for Classification ?

• ROC Curve

• Precision Recall curve

• Histogram of residual

• Predict vs True value