Comparison Of Support Vector Machine And Naïve Bayes Algorithms For Analyzing Public Interest In Espresso Coffee

Abstract

Given its increasing popularity, public interest in buying espresso coffee is an important concern for coffee industry players. To understand and predict this buying interest, the use of classification algorithms in data analysis is crucial. This study was conducted to compare the performance of two popular classification algorithms, namely Support Vector Machine and Naïve Bayes, in analyzing public interest in buying espresso coffee. This research problem is based on the need for an accurate predictive model in the coffee industry to aid in strategic decision-making related to marketing and sales. The proposed solution is to implement two different classification algorithms and assess their performance using a variety of performance evaluation metrics. The purpose of this study is to determine which algorithm is superior in terms of accuracy, precision, recall, and f1-score. The research method entails collecting data on public interest in purchasing espresso coffee, preprocessing data, implementing both algorithms, and evaluating each algorithm's performance. The results show that Naïve Bayes consistently outperforms Support Vector Machine in all performance evaluation metrics. Naïve Bayes achieved 94.00% accuracy, 91.40% precision, 100% recall, and 95.51% F1-Score, compared to Support Vector Machine, which achieved 90.00% accuracy, 88.60% precision, 96.90% recall, and 92.56% F1-Score. The conclusion of this study is that the Naïve Bayes classifier is more effective and efficient in predicting people's purchasing interest in espresso coffee compared to support vector machines. This advantage can be attributed to the ability of Naïve Bayes to handle data that may have non-normal distributions or independent variables.