Ink Wardana, Dyah K. Agustika, Setio Basuki
This paper aims to develop a machine learning model to classify herbal teas using embedded machine learning. Classifying the herbal teas is important due to considering its popularity as a natural remedy approach. Currently, there are several studies focusing on Electronic Nose (E-Nose) as a main technique to capture gasses emitted by herbal teas. This paper develops a machine learning model to be applied in devices with limited power, memory, and processing capacity. The model is used to classify four types of herbal teas namely Chamomile, Lemon Ginger, Peppermint, and Red Berries. To realize this, this paper uses Deep Learning implemented by using TinyML as an embedded machine learning platform. The gasses emitted by herbal teas are captured by Raspberry Pi Pico-based controller, preprocessed as a tabular dataset, then forwarded to machine learning model development. The experiments were conducted on 2,400 samples; this paper obtained the result that the model achieves an accuracy of 92.33% in classifying the herbal tea samples. © 2025 Author(s).
Department of Electrical Engineering, Politeknik Negeri Bali, Badung, Indonesia; Department of Physics Education, Universitas Negeri Yogyakarta, Yogyakarta, Indonesia; Department of Informatics, Universitas Muhammadiyah Malang, Malang, Indonesia; School of Engineering, University of Warwick, Coventry, United Kingdom
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock