Logistic Regression & Binary Classification

Which loss function is typically used in binary classification problems?

• Mean Squared Error Loss

• Binary Cross Entropy Loss

• Categorical Cross Entropy Loss

• Huber Loss

What does the Gradient Descent algorithm primarily aim to minimize?

• The classification error rate

• The model complexity

• The cost function

• The number of features

What role does the learning rate play in Gradient Descent?

• It determines the size of the steps taken towards the minimum of the cost function

• It specifies the initial value of the weights

• It controls the number of iterations in the training process

• It adjusts the bias values of the model

Which metric is NOT directly used to evaluate binary classification models?

• Accuracy

• Recall

• Precision

• Mean Absolute Error

Which statement best describes the relationship between Sensitivity and Specificity in a classifier?

• They are directly proportional to each other

• They are inversely proportional to each other

• Improving one necessarily worsens the other

• There is no direct relationship; they measure different aspects of the classifier's performance

What is the primary advantage of using K-Fold Cross Validation in model evaluation?

• It simplifies the model training process

• It increases the speed of the evaluation

• It ensures that every data point is used for both training and testing

• It reduces the need for a validation set

What is the primary goal of binary classification?

• To predict a continuous value

• To classify data into two distinct categories

• To cluster data into multiple groups

• None of the above

Please define the Sigmoid function

• (1+e^-x)²

• e/(1+e^x)

• 1/(1+e^-x)

• (1+e)/(1-e^x)

Why we can't use linear regression for binary classification?

• Because the linear function can only output the real numbers

• Because the data can't be separated by using single line

• Because there are only two possible true labels (1 or 0)

• Because we can't find optimum value

What does the area under the ROC curve (AUC) indicate?

• The model's ability to handle unbalanced datasets

• The model's capacity to distinguish between the classes

• The precision of the model

• The recall of the model