Coffee Aroma Classification Based on an Electronic Nose: Comparative Evaluation of DNN, SVM, and ANN with IQR and Wavelet Preprocessing
DOI:
https://doi.org/10.30871/jaic.v10i2.12207Keywords:
Electronic Nose, Coffee Aroma Classification, Deep Neural Network, Support Vector Machine, Artificial Neural Network, Discrete Wavelet TransformAbstract
Coffee aroma assessment is commonly performed through cupping tests, which are subjective and highly dependent on human perception. This study proposes an objective approach for multi-class coffee aroma classification using a low-cost Electronic Nose (E-Nose) and a comparative evaluation of Deep Neural Network (DNN), Support Vector Machine (SVM), and Artificial Neural Network (ANN). Aroma data were acquired from an ESP32-based sensor array consisting of MQ-9, MQ-135, and DHT22, resulting in 1,503 samples from 10 Arabica and Robusta coffee classes. Two preprocessing scenarios were evaluated: (1) raw data + Interquartile Range (IQR) + Discrete Wavelet Transform (DWT) + Min–Max scaling, and (2) raw data + Min–Max scaling. The models were evaluated using an 80:20 train–test split and 5-fold cross-validation on the training set. The results show that DNN achieved the best performance in both scenarios, reaching 91.45% accuracy in the first scenario and 80.07% in the second. SVM also improved substantially from 61.46% to 81.78% after applying IQR and DWT. In contrast, ANN performed better in the second scenario, achieving 73.72% accuracy compared to 61.43% in the first. These findings indicate that preprocessing effectiveness depends on the classification model, while DNN remains the most robust and consistent model for multi-class coffee aroma classification based on E-Nose tabular data.
Downloads
References
[1] K. Agropolitan and K. Kudus, “Pengembangan Potensi Kopi Sebagai Komoditas Unggulan Kawasan Agropolitan Kabupaten Kudus,” vol. 21, 2024.
[2] F. Amalia et al., “Identification of potential quality markers in Indonesia’s Arabica specialty coffee using GC/MS-based metabolomics approach,” Metabolomics, vol. 19, no. 11, pp. 1–11, 2023, doi: 10.1007/s11306-023-02051-5.
[3] C. Gonzalez Viejo, E. Tongson, and S. Fuentes, “Integrating a low‐cost electronic nose and machine learning modelling to assess coffee aroma profile and intensity,” Sensors, vol. 21, no. 6, pp. 1–16, 2021, doi: 10.3390/s21062016.
[4] C. H. Lee, “An AI-powered Electronic Nose System with Fingerprint Extraction for Aroma Recognition of Coffee Beans,” Sensors, 2022.
[5] M. Telaumbanua, D. D. Novita, S. Triyono, and C. Saragih, “Tipe Chamber Dan Posisi Sensor E-Nose Untuk Mendeteksi Aroma Biji Kopi Robusta Menggunakan Mikrokontroler,” J. Ilm. Rekayasa Pertan. dan Biosist., vol. 9, no. 1, pp. 84–95, 2021, doi: 10.29303/jrpb.v9i1.237.
[6] D. P. Purbawa et al., “Adaptive filter for detection outlier data on electronic nose signal,” Sens. Bio-Sensing Res., vol. 36, no. April, p. 100492, 2022, doi: 10.1016/j.sbsr.2022.100492.
[7] A. S. AlSalehy and M. Bailey, “Improving Time Series Data Quality: Identifying Outliers and Handling Missing Values in a Multilocation Gas and Weather Dataset,” Smart Cities, vol. 8, no. 3, pp. 1–39, 2025, doi: 10.3390/smartcities8030082.
[8] F. G. Handayani, A. Rajagukguk, I. H. Rosma, and E. Ervianto, “Comparative Analysis Of Mother Wavelet For Voltage Sag And Swell Classification Using Discrete Wavelet Transform ( Dwt ) And Radial Basis Function Neural Network ( Rbfnn ) Analisis Komparatif Mother Wavelet Untuk Klasifikasi Gangguan Voltage Sag Dan Swell,” vol. 25, no. 2, pp. 132–140, 2025.
[9] M. A. Barata et al., “Hydrogen Sulfide Leak Detection Using The C4.5 Algorithm: Optimizing Feature Extraction For Enhanced Accuracy Mula,” vol. 1, pp. 348–358, 2024.
[10] I. Wisma, D. Prastya, and R. Sarno, “Coffee Aroma Classification Based On World Coffee Research Sensory Lexicon Using Electronic Nose,” no. March, pp. 19–23, 2025, doi: 10.31602/.v0i0.15389.
[11] A. Syafi’i, M. A. Barata, and R. Rohmah, “Artificial Neural Network Algorithm on e-Nose Devices for Honey Classification,” J. Telemat., vol. 20, no. 1, pp. 28–48, 2025, doi: 10.61769/telematika.v20i1.722.
[12] E. Rianty and K. Budi, “Perbandingan Kinerja Algoritma C4 . 5 dan Naive Bayes Dalam Klasifikasi Data Penjualan Buku PT . XYZ,” vol. 12, no. 6, 2025, doi: 10.30865/jurikom.v12i6.9345.
[13] Malikhah et al., “Detection of Infectious Respiratory Disease Through Sweat From Axillary Using an E-Nose With Stacked Deep Neural Network,” IEEE Access, vol. 10, pp. 51285–51298, 2022, doi: 10.1109/ACCESS.2022.3173736.
[14] D. Erwanto et al., “Development and Application of an Electronic Nose System for Classifying Coffee Varieties Based on Aromatic Profiles,” J. Intell. Syst. Control, vol. 3, no. 3, pp. 186–200, 2024, doi: 10.56578/jisc030305.
[15] S. D. Astuti et al., “Electronic nose coupled with artificial neural network for classifying of coffee roasting profile,” Sens. Bio-Sensing Res., vol. 43, no. February, p. 100632, 2024, doi: 10.1016/j.sbsr.2024.100632.
[16] V. Borisov, T. Leemann, K. Seßler, J. Haug, M. Pawelczyk, and G. Kasneci, “Deep Neural Networks and Tabular Data : A Survey,” no. June, pp. 1–22, 2022.
[17] F. Haselzadeh, “E-nose equipped with Artificial Intelligence Technology for diagnosis of dairy cattle disease in veterinary,” 2021.
[18] D. B. Magfira, F. Yudianto, T. D. Wulan, T. Herlambang, R. P. N. Budiarti, and A. T. R. Siswanti, “Perancangan IoT Sederhana Untuk Sistem Pendeteksi Kemurnian Kopi Bubuk,” J. Tecnoscienza, vol. 9, no. 1, pp. 86–96, 2024, doi: 10.51158/cykvay82.
[19] B. Sumanto, D. R. Java, W. Wijaya, and J. Hendry, “Seleksi Fitur Terhadap Performa Kinerja Sistem E-Nose untuk Klasifikasi Aroma Kopi Gayo,” MATRIK J. Manajemen, Tek. Inform. dan Rekayasa Komput., vol. 21, no. 2, pp. 429–438, 2022, doi: 10.30812/matrik.v21i2.1495.
[20] I. S. Nasution, D. P. Delima, Z. Zaidiyah, and R. Fadhil, “Mathematical Modelling of Engineering Problems A Low Cost Electronic Nose System for Classification of Gayo Arabica Coffee Roasting Levels Using Stepwise Linear Discriminant and K-Nearest Neighbor,” vol. 9, no. 5, pp. 1271–1276, 2022.
[21] N. Dzakwan, S. Gandhi, M. Y. Hariyawan, and H. Briantoro, “Stasiun Cuaca Mini Pintar Berbasis Deep Learning,” vol. 12, no. 5, pp. 8026–8033, 2025.
[22] M. J. Vikri, I. W. D. Prastya, U. P. Sanjaya, and M. A. Barata, “Rice Quality Identification for Indonesian Food Standards Based on Electronic Nose Identifikasi Kualitas Beras Berdasarkan Standar Pangan Indonesia Berbasis Electronic Nose,” no. 1, 2025.
[23] P. P. Allorerung, A. Erna, M. Bagussahrir, and S. Alam, “Analisis Performa Normalisasi Data untuk Klasifikasi K-Nearest Neighbor pada Dataset Penyakit,” JISKA (Jurnal Inform. Sunan Kalijaga), vol. 9, no. 3, pp. 178–191, 2024, doi: 10.14421/jiska.2024.9.3.178-191.
[24] M. A. Barata, E. Noersasongko, Purwanto, and M. A. Soeleman, “Improving the Accuracy of C4.5 Algorithm with Chi-Square Method on Pure Tea Classification Using Electronic Nose,” vol. 5, no. 158, pp. 226–235, 2026.
[25] F. I. Silfana and M. A. Barata, “Using K-NN Algorithm for Evaluating Feature Selection on High,” vol. 17, no. 2, 2024.
[26] G. Zeng, “Invariance Properties and Evaluation Metrics Derived from the Confusion Matrix in Multiclass Classification,” Mathematics, vol. 13, no. 16, 2025, doi: 10.3390/math13162609.
[27] Z. Peng et al., “A Comprehensive Evaluation Model for Optimizing the Sensor Array of Electronic Nose,” Appl. Sci., vol. 13, no. 4, 2023, doi: 10.3390/app13042338.
[28] J. Tampubolon, “Comparative performance of LSTM and DNN in sentiment analysis,” vol. 14, no. 1, pp. 1–11, 2025, [Online]. Available: www.ejournal.isha.or.id/index.php/Mandiri
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Muhammad Yulvi Aditya Pradana, Ifnu Wisma Dwi Prasetya, Ita Aristia Sa’ida
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License (Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) ) that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
