saniyaabushakimova / red-wine-quality-classification-using-supervised-learning

Conducted data analysis using statistical tools and complex visualizations; trained logistic regression, k-nn, kernelized svm, and random forest models; performed hyperparameter tuning and error analysis. Tech: Python (Seaborn, Matplotlib, Pandas, Scikit-Learn)

Project_1

Project Type: Exploratory Data Analysis

Purpose

• Understand the relationship between the quality of red wine and its chemical properties
• Answer the following research questions:
  • What is the difference between low-quality and high-quality red wine in mean acidity, which include fixed acidity, volatile acidity, citric acid, and pH?
  • Does alcohol have a positive correlation with red wine quality?
  • What is the proportion of high-quality red wines whose alcohol concentration is larger than the average? What is the proportion of low-quality red wines whose alcohol concentration is lower than the average?

Content

• Project_1/Project_1_solution.ipynb- the project report and a code written in Python.

Project_2

Project Type: Classification problem

Purpose

• Construct a classification task that will predict the quality of red wine based on its chemical properties.

Machine Learning tools

• Selected models (Logistic Regression, K-NN, Kernelized SVM, Random Forest);
• Evaluation Metric (Accuracy, F1, Precision, Recall);
• Hyperparameter tuning;
• Checking for overfitting;
• Error Analysis (ROC curves, Confusion Matrix, Feature Importance).

Content

• Project_2/Project_2_solution.ipynb- the project report and a code written in Python.