(Solved): Jerome Tang asked his analytics team to help him model the probability of his team Winning a basket ...

Jerome Tang asked his analytics team to help him model the probability of his team Winning a basketball gane. Here. Winning is a 0 . -1 valued variable where Winning $=1$ denotes that his team won. A analyst chooses to use logistie regression to model this probablity. using the number of Assists and the number of Turnovers as predictors. That is, the analyst models the probabily of winning a basketball game as follows: $P(Winning=1)=\exp ({?}_0+{?}_1(Assists)+{?}_2$ (Turnovers $)y(1+\exp ({?}_0+?+(Assists)+{?}_2($ turnovers $):$ where $\exp (x)$ denotes $e^x$. The parameter estimates of the logistic regression modiel are as follows. a. (1pt) From these results, do more turnovers result in a higher or lower chance of winning a basketball game? Higher, since the Intercept term is negative. Higher, since the Assists term is positive. Higher, since the Turnovers term is positlve. Lower, since the Intercept term is negative. Lower, since the Assists term is positive. Lower, since the Turnovers term is positive. b. (2pts) What is the predicted probablitv of Winnina implied bv the loaistic model for when the team has 7 assists and 13 turnovers? b. (2pts) What is the predicted probability of Winning implied by the logistic modet for when the team has 7 assists and 13 turnovers? Note: Round all answers to at least 4 decimal places. Report the probability between 0 and 1 (not in \%). c. (1pt) Another analyst on the team chooses to model the probability usibg a linear probability moder instead. Which of the following are true about the linear probability model? Since the response is 0 - 1 valued, the linear probability model may not minimize the sums of squares of the residuals. Standard errors from typical $R$ output using the linear probability model are accurate. Because the linear probability model is estimating probabilities, the predicted probability of winning given by this model will always be between 0 and 1. The linear probability model is the most widely used binary choice model in practice. Coefficients for a linear probability model are hard to interpret None of the above are true about the linear probability model d. (1pt) Jerome Tang is debating whether to use the logistic regression model or the linear probability model for future analyses. Which of the following measures is most appropriate when selecting the best binary cholce model? Mean square error (MSE) of the responses. R-Squared F-Statistic The heteroskedasticity coefficient T-Statistic Adjusted R-Squared Prediction accuracy