Recall the equation for logistic regression: 1 P(ya) = Itewr In this equation the output of a linear function is fed through the S-shaped "logistic" or ("sigmoid") function, which maps real values to the range [0,1]. 1 Is the decision boundary for a logistic regression model linear? 2 (a) Given x = [1, 5], wix = -1.1 + 3x1 - 0.7x2, what is the predicted label y for it? (b) Given the above data point and model, if we have another sample a' = [3, 1], please compute the log-likelihood for them. 3. Applying the logistic regression model to a binary classification task, it returns us predictions (0, 1,1,0,1) while the true values are (0,0,1,1,1). Please compute accuracy, precision, and recall for the model performance. 4. Suppose we would like to use this model to design a system to screen out a rare abnormality in chest x-rays. Which of the following ones is the most important factor to consider when we design the system, specificity, precision, sensitivity, or accuracy? Please give your reason