2.The sigmoid function always outputs a value in which range?
3.What is the output of the sigmoid function when z = 0?
4.In logistic regression with threshold t = 0.5, the decision boundary is defined by which condition?
5.Why is Mean Squared Error (MSE) not used as the cost function for logistic regression?
6.For a training example with true label y = 1 and predicted probability p-hat = 0.05 what is the nature of the log-loss contribution?
7.The gradient of log-loss with respect to theta_j in logistic regression is:
8.In One-vs-Rest multiclass classification with K = 4 classes how many binary classifiers are trained?
9.Which statement about the softmax function is correct?
10.A spam filter flags 100 emails as spam. Of these 80 are truly spam and 20 are legitimate. There are also 50 spam emails it missed. What is the precision?
11.Recall is the most important metric in which scenario?
12.The ROC curve plots True Positive Rate on the y-axis against which quantity on the x-axis?
13.An AUC of 0.5 indicates:
14.In scikit-learn's LogisticRegression the parameter C = 0.01 compared to C = 100 represents:
15.A logistic regression model trained where classes are perfectly linearly separable is likely to exhibit: