Using Statsmodels, load the American National Election Studies dataset and train a Multinomial Logistic Regression model...

Transcribed Image Text:

Using Statsmodels, load the American National Election Studies dataset and train a Multinomial Logistic Regression model (MNLogit). You can load the dataset as follows: import statsmodels. api as sm anes_data =sm.datasets. anes96.load() X = sm.add_constant (anes_data. exog, prepend=False) y = anes_data. endog (a) (3 points) Explain why the summary shows multiple values for each coefficient. (b) (2 points) Looking at the p-values, are any of the features insignificant? (c) (2 points) Are there any features that are useful for only few of the classes? If so, list the features and corresponding classes. (d) (2 points) Are there any features that are useful for all the classes? If so, list the features. Using Statsmodels, load the American National Election Studies dataset and train a Multinomial Logistic Regression model (MNLogit). You can load the dataset as follows: import statsmodels. api as sm anes_data =sm.datasets. anes96.load() X = sm.add_constant (anes_data. exog, prepend=False) y = anes_data. endog (a) (3 points) Explain why the summary shows multiple values for each coefficient. (b) (2 points) Looking at the p-values, are any of the features insignificant? (c) (2 points) Are there any features that are useful for only few of the classes? If so, list the features and corresponding classes. (d) (2 points) Are there any features that are useful for all the classes? If so, list the features.