Dataset

The data has been acquired from funding hence cannot be shared but to understand the code this is the head of the data frame.

Brand Name is used as index and MA as the dependent variable. Other columns indicate the independent variables.

Logistic Model

When the dependent variable is dichotomous, logistic regression is the only regression analysis to use (binary). The logistic regression, like all regression analyses, is a statistical analysis. To characterise data and illustrate the relationship between one dependent binary variable and one or more nominal, ordinal, interval, or ratio-level independent variables, logistic regression is used.

Map to Project

• Initially the data is cleared by removing missing values.
• Splitting of the Dataset With scikit learn's (sklearn) train_test_split() into training and test dataset for unbiased evaluation.
• Using sklearn to fit the trainng model on the data set.
• Checking accuracy of the model with accuracy score(sklearn.metrics), confusion matrix and ROC curve.
• Performing logistic regression using statsmodel.