A study on the quality of wine(red and white) using Machine Learning models.
Using the Wine Quality Dataset this report will focus on the following problem set:
- Illustrate the following summary information in Python.
a. Read the data and construct some appropriate graphs for each of the variables with interpretation.
b. Conduct the EDA for the selected study data.
c. Construct the pair plots for different wine qualities (Good >5, Bad<=5) in Wine Quality Data.
-
Classify the different wine qualities (Good >5, Bad<=5) in Wine Quality Data. Hence, estimate the Logistic Regression model to predict the probability of good wine quality in Wine Quality Data. Interpret the fitted model.
-
Split the entire dataset into the training dataset (70%) and the test dataset (30%). To classify the different wine qualities (Good >5, Bad<=5) in Wine Quality Data use the following supervised learning methods for the training dataset:
a. Run the Logistic Regression model to classify the different categories. Find the confusion matrix and ROC curve. Hence, calculate and interpret the predictive value positive and negative, Accuracy, Sensitivity, and Specificity of the test.
b. Run the Decision Trees to classify the different categories. Find the confusion matrix and ROC curve. Hence, calculate and interpret the predictive value positive and negative, Accuracy, Sensitivity, and Specificity of the test.
c. Run the Random Forest to classify the different categories. Find the confusion matrix and ROC curve. Hence, calculate and interpret the predictive value positive and negative, Accuracy, Sensitivity, and Specificity of the test.
d. Run the Support Vector Machines to classify the different categories. Find the confusion matrix and ROC curve. Hence, calculate and interpret the predictive value positive and negative, Accuracy, Sensitivity, and Specificity of the test.
e. Hence, evaluate the best-performed model for this study using the test dataset.