Wine Origin Classification 🍷

This repository is dedicated for the jupyter notebook that used to perform classification of wines' origin based on its chemical composition. This open dataset is available on UC Irvine Machine Learning Repository which can be publicly accessible using the python module named ucimlrepo. It can be installed using comand below

pip install ucimlrepo

Algorithm used in the classification 💻

We adopted supervised learning for this classification task, primarily look into the DecisionTreeClassifier and RandomForestClassifier offered by the scikit learn module in python

How's was the performance of algorithm? 📈

• DecisionTreeClassifier

The best decision tree was shown as above with maximum depth of 4 and accuracy score of 0.97 for the test set. You may further refer to the Decision_tree.ipynb file.

• RandomForestClassifier

The figure above show one of the tree in best estimators from the GridSearchCV using the Random Forest algorithm. The best estimator was with the maximum depth of 10, 75 trees and random state of 42. The accuracy score was 1.0 for the test set. You may further refer to the RDF_WInes.ipynb file.

Built With

• jupyter-notebook