A Proposed Classification Method in Menu Engineering Using the K-Nearest Neighbors Algorithm

Abstract

In the culinary business, the menu is crucial; therefore, the performance of each menu needs to be known to maintain business continuity. Menu engineering is a special technique used to see the performance comparison of each menu item. This research proposes modeling menu engineering with a new approach in classifying menu items using the k-Nearest Neighbors (k-NN) algorithm using the sales training data of sales data in 2019 belonging to one of the micro, small and medium-sized enterprises in the culinary subsector in Salatiga, Indonesia. In the modeling, the popularity index (menu mix) and item contribution margin are used as variables, while the menu item class is used as the label attribute of the classification. Determination of the k value in the k-NN algorithm was done by the experimental method so that it produces the most optimal k based on the highest accuracy value, while the distance calculation on k-NN was done using euclidean distance. Evaluation of the model was done using 10-fold cross-validation with four performance evaluation criteria, namely weighted mean recall, weighted mean precision, accuracy, classification error. Based on the evaluation results, an accuracy of 96.84% was obtained; thus, the proposed model is considered to have given good and accurate results. This proposed model has been implemented in MSME sales data to classify menu items. The results of this classification were used as a basis for recommending menu engineering strategies to MSMEs.

Comparison of K-Means & K-Means++ Clustering Models using Singular Value Decomposition (SVD) in Menu Engineering

“…The cluster output is used to create strategic recommendations for the menu to increase sales. Setiyawati [1] also proposed an engineering menu model using the K-Nearest Neighbors Algorithm. The findings of this study suggested a new approach to engineering menu analysis using a classification of menu items according to the characteristics of the menu mix and item contribution margin.…”

“…The menu is one of the most fundamental aspects of business continuity in the culinary industry [1], [2]. A welldesigned and managed menu could generate greater profit for the business and provide product information to consumers [2], [3].…”

The combination of color-texture features and machine learning for detecting Dayak beads

“…Commonly used texture features include Local Binary Pattern [12]- [14] and GLCM [6], [15], [16]. Furthermore, classification frequently applied using the method of KNN [12], [17], ANN [18], [19], SVM [20], [21], and Naïve Bayes [22], [23].…”