Optimized higher-order photon state classification by machine learning DOI Creative Commons
Guangpeng Xu, Jeffrey Carvalho, Chiran Wijesundara

и другие.

Deleted Journal, Год журнала: 2024, Номер 1(3)

Опубликована: Сен. 1, 2024

The classification of higher-order photon emission becomes important with more methods being developed for deterministic multiphoton generation. widely used second-order correlation g(2) is not sufficient to determine the quantum purity higher Fock states. Traditional characterization require a large amount detection events, which leads increased measurement and computation time. Here, we demonstrate machine learning model based on 2D Convolutional Neural Network (CNN) rapid states up |3⟩ an overall accuracy 94%. By fitting g(3) simulated exhibits efficient performance particularly sparse data, 800 co-detection events achieve 90%. Using proposed experimental setup, this CNN classifier opens possibility quasi-real-time states, holds broad applications in technologies.

Язык: Английский

Optical properties estimation of photonic crystal fiber using Gaussian process regression DOI Creative Commons

Sk Md Abdul Kaium Kaium,

Md. Aslam Mollah

Optics Continuum, Год журнала: 2024, Номер 3(8), С. 1369 - 1369

Опубликована: Июль 5, 2024

In contrast to typical optical fiber, photonic crystal fiber (PCF) exhibits a variety of unique properties as result its flexible cladding distribution. Nonetheless, assessing PCF characteristics becomes difficult when structural parameters fluctuate. This issue is serious impediment fully understanding and leveraging PCF's potential for diverse applications. Furthermore, the in factors makes it ensure consistent reliable performance practical systems. Artificial neural networks (ANN) are widely used forecast PCF. However, ANNs have issues dealing with local minima. contrast, solutions obtained from support vector machines regressions (SVM/SVR), Gaussian process (GPR), k-nearest neighbors regression (KNNR) globally avoid dangers slipping into minimum values. Major such effective refractive index ( n e f ), confinement loss α c ) dispersion D were predicted using SVM/SVR, GPR, KNNR, random forest (RFR), gradient boosting (GBR), ANN. To evaluate various algorithms, we created database 2912 samples including X Y directions. terms prediction accuracy stability, SVM GPR outperform other approaches.

Язык: Английский

Процитировано

1
Optimized higher-order photon state classification by machine learning DOI Creative Commons
Guangpeng Xu, Jeffrey Carvalho, Chiran Wijesundara

и другие.

Deleted Journal, Год журнала: 2024, Номер 1(3)

Опубликована: Сен. 1, 2024

The classification of higher-order photon emission becomes important with more methods being developed for deterministic multiphoton generation. widely used second-order correlation g(2) is not sufficient to determine the quantum purity higher Fock states. Traditional characterization require a large amount detection events, which leads increased measurement and computation time. Here, we demonstrate machine learning model based on 2D Convolutional Neural Network (CNN) rapid states up |3⟩ an overall accuracy 94%. By fitting g(3) simulated exhibits efficient performance particularly sparse data, 800 co-detection events achieve 90%. Using proposed experimental setup, this CNN classifier opens possibility quasi-real-time states, holds broad applications in technologies.

Язык: Английский

Процитировано

0