Enhancing the Performance of Photonic Sensor Using Machine-Learning Approach
Yogendra Swaroop Dwivedi, Rishav Singh, Anuj K. Sharma, Ajay Kumar Sharma
Abstract
This article reports on the implementation of adequate machine-learning (ML) models on different datasets vis-a-vis fiber-optic plasmonic sensor devices. The variation of the sensor’s figure of merit (FOM) with light wavelength ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\lambda $ </tex-math></inline-formula> ) and metal layer thickness ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${d}_{m}$ </tex-math></inline-formula> ) is considered as a starting point and accordingly, the appropriate ML model is chosen. The FOM datasets were found to be consistent with the Gaussian process regressor (GPR) model. The application of GPR with finer resolution (0.001 nm) of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\lambda $ </tex-math></inline-formula> on the datasets led to enhanced magnitudes of the sensor’s FOM. The dataset (459 points) having nine different values of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${d}_{m}$ </tex-math></inline-formula> led to a predicted FOM of 6526.23 at <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\lambda =1099.343$ </tex-math></inline-formula> nm. Furthermore, the dataset (714 points) having 13 different values of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${d}_{m}$ </tex-math></inline-formula> led to a predicted FOM value of 6356.98 at <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\lambda =1099.345$ </tex-math></inline-formula> nm. These are promising results as far as the application of the sensor in biosensing is concerned. Furthermore, the chosen model is found to be highly consistent with the data in terms of trend matching, and the values of other evaluation parameters [e.g., <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}^{\,{2}}$ </tex-math></inline-formula> and mean absolute error (MAE)] are found to be in considerably desirable ranges. This study clearly reveals that the selection of an appropriate ML model and its implementation on various datasets can lead to more efficient finalization of the sensor design with enhanced sensing performance. This process is critical before the actual experimental realization of the finalized sensor design.