Cuisine Prediction from Ingredients using Hyper Parameter Tuning on Machine Learning Algorithms

This study focuses on analyzing the correlation between cuisines and ingredients by employing different classification techniques of machine learning. Utilizing a Kaggle dataset, we compare the accuracy of multiple classifiers by applying linear support vector classifier(Linear SVC), extreme gradient boosting (XGBoost), decision trees, random forest, logistic regression, k-nearest neighbor, and naive Bayes. The LSVC achieves an accuracy of 80.43%. Applying the hyper-parameter tuning, the accuracy is further improved, which is 81.62%. This enhancement carries significant implications, as even small accuracy improvement can provide more reliable predictions and valuable insights for understanding cuisine and ingredient patterns in a large-scale dataset.