Presentation Information
[11a-PA3-8]Machine learning-based free chlorine sensor for buffering-induced current variations
〇(M1)Koki Anzai1, Mayano Yamanouchi1, Takeshi Watanabe1, Yasufumi Yokoshiki1, Isao Hiyama2, Nobuyuki Sanari2, Shinji Koh1 (1.Aoyama Gakuin Univ., 2.Shibata Scientific Technology LTD.)
Keywords:
Electrochemistry,Machine Learning
To achieve highly accurate measurement of free chlorine concentration, which is essential for hygiene management, I developed a sensor that combines electrochemical measurement with machine learning. Conventionally, training data were collected using buffer solutions; however, differences in sensor responses between buffer solutions and real-world samples limited the estimation accuracy. To address this issue, I generated training data using non-buffered solutions and investigated their effectiveness in improving concentration estimation accuracy for real samples.