Data-driven gradient algorithm for high-precision quantum control

Food waste and food insecurity are two challenges that coexist in many communities. To mitigate the problem, food rescue platforms match excess food with the communities in need, and leverage external volunteers to transport the food. However, the external volunteers bring significant uncertainty to the food rescue operation. We work with a large food rescue organization to predict the uncertainty and furthermore to find ways to reduce the human dispatcher's workload and the redundant notifications sent to volunteers. We make two main contributions. (1) We train a stacking model which predicts whether a rescue will be claimed with high precision and AUC. This model can help the dispatcher better plan for backup options and alleviate their uncertainty. (2) We develop a data-driven optimization algorithm to compute the optimal intervention and notification scheme. The algorithm uses a novel counterfactual data generation approach and the branch and bound framework. Our result reduces the number of notifications and interventions required in the food rescue operation. We are working with the organization to deploy our results in the near future.

Download Full-text

Control Method for PEMFC Using Improved Deep Deterministic Policy Gradient Algorithm

Frontiers in Energy Research ◽

10.3389/fenrg.2021.753064 ◽

2021 ◽

Vol 9 ◽

Author(s):

Jiawen Li ◽

Yaping Li ◽

Tao Yu

Keyword(s):

Optimal Control ◽

Output Voltage ◽

Control Method ◽

Voltage Control ◽

Control Policy ◽

Gradient Algorithm ◽

Data Driven ◽

Q Learning ◽

Policy Gradient

A data-driven PEMFC output voltage control method is proposed. Moreover, an Improved deep deterministic policy gradient algorithm is proposed for this method. The algorithm introduces three techniques: Clipped multiple Q-learning, policy delay update, and policy smoothing to improve the robustness of the control policy. In this algorithm, the hydrogen controller is treated as an agent, which is pre-trained to fully interact with the environment and obtain the optimal control policy. The effectiveness of the proposed algorithm is demonstrated experimentally.

Download Full-text

A Modified Deep Deterministic Policy Gradient Algorithm for Data-Driven Inventory Management

Journal of the Korean Society of Supply Chain Management ◽

10.25052/kscm.2021.12.21.3.71 ◽

2021 ◽

Vol 21 (3) ◽

pp. 71-89

Author(s):

Byeongkwon Lee ◽

Kun-Soo Park ◽

Se-Youn Jung

Keyword(s):

Inventory Management ◽

Gradient Algorithm ◽

Data Driven ◽

Policy Gradient

Download Full-text

Research on a Novel Compensation Algorithm for MEMS Sub-Pixel Displacement Measurement

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.901.81 ◽

2014 ◽

Vol 901 ◽

pp. 81-86

Author(s):

Zhang Fang Hu ◽

Dong Dong Huang ◽

Yuan Luo ◽

Yi Zhang

Keyword(s):

High Precision ◽

Order Term ◽

Micro Electro Mechanical System ◽

Gradient Algorithm ◽

Taylor Formula ◽

Noise Resistance ◽

High Precision Measurement ◽

Key Technology ◽

Compensation Algorithm ◽

Higher Order Term

The sub-pixel allocation is a key technology for achieving high-precision measurement for micro electro mechanical system (MEMS). In this paper, a novel sub-pixel compensation algorithm based on an improved gradient algorithm utilized in a rough sub-pixel position is proposed to compensate the insufficient accuracy extracted by the surface fitting. And compared with traditional gradient used in the integer pixel, the proposed algorithm can reduce the error introduced by abandoning the higher order term without using iteration and second-order Taylor formula. The experimental results show that the displacement parameters calculated by the proposed algorithm is more accurate, and the method has a good noise resistance, it can meet high-precision positioning of MEMS motion image.

Download Full-text

Gradient algorithm applied to laboratory quantum control

Physical Review A ◽

10.1103/physreva.79.053417 ◽

2009 ◽

Vol 79 (5) ◽

Cited By ~ 40

Author(s):

Jonathan Roslund ◽

Herschel Rabitz

Keyword(s):

Quantum Control ◽

Gradient Algorithm

Download Full-text

An Unsupervised Data-Driven Cross-Lingual Method for Building High Precision Sentiment Lexicons

2013 IEEE Seventh International Conference on Semantic Computing ◽

10.1109/icsc.2013.40 ◽

2013 ◽

Cited By ~ 1

Author(s):

Pierluca Sangiorgi ◽

Agnese Augello ◽

Giovanni Pilato

Keyword(s):

High Precision ◽

Data Driven ◽

Cross Lingual

Download Full-text

Reprogrammable and high-precision holographic optical addressing of trapped ions for scalable quantum control

npj Quantum Information ◽

10.1038/s41534-021-00396-0 ◽

2021 ◽

Vol 7 (1) ◽

Author(s):

Chung-You Shih ◽

Sainath Motlakunta ◽

Nikhil Kotibhaskar ◽

Manas Sajjan ◽

Roland Hablützel ◽

...

Keyword(s):

High Precision ◽

Quantum Control ◽

Quantum Information Processing ◽

Trapped Ions ◽

Optical Manipulation ◽

High Susceptibility ◽

Ion Separation ◽

Quantum Particles ◽

Ion Species

AbstractHigh-precision, individually programmable manipulation of quantum particles is crucial for scaling up quantum information processing (QIP) systems such as laser-cooled trapped-ions. However, restricting undesirable “crosstalk” in optical manipulation of ion qubits is fundamentally challenging due to micron-level inter-ion separation. Further, inhomogeneous ion spacing and high susceptibility to aberrations at UV wavelengths suitable for most ion-species pose severe challenges. Here, we demonstrate high-precision individual addressing (λ = 369.5 nm) of Yb+ using a reprogrammable Fourier hologram. The precision is achieved through in-situ aberration characterization via the trapped ion, and compensating (to λ/20) with the hologram. Using an iterative Fourier transformation algorithm (IFTA), we demonstrate an ultra-low (<10−4) intensity crosstalk error in creating arbitrary pair-wise addressing profiles, suitable for over fifty ions. This scheme relies on standard commercial hardware, can be readily extended to over a hundred ions, and adapted to other ion-species and quantum platforms.

Download Full-text

Inversion of controlled-source electromagnetic reflection responses

Geophysics ◽

10.1190/geo2015-0320.1 ◽

2016 ◽

Vol 81 (5) ◽

pp. F49-F57 ◽

Cited By ~ 6

Author(s):

Jürg Hunziker ◽

Jan Thorbecke ◽

Joeri Brackenhoff ◽

Evert Slob

Keyword(s):

Electromagnetic Fields ◽

Synthetic Data ◽

Water Layer ◽

Gradient Algorithm ◽

Data Driven ◽

Data Sets ◽

Controlled Source ◽

Material Parameters ◽

Nonlinear Conjugate Gradient ◽

Position And Orientation

Marine controlled-source electromagnetic reflection responses can be retrieved by interferometry. These reflection responses are free of effects related to the water layer and the air above it and do not suffer from uncertainties related to the source position and orientation. Interferometry is a data-driven process requiring proper sampling of the electromagnetic field as well as knowledge of the material parameters at the receiver level, i.e., the sediment just below the receivers. We have inverted synthetic data sets using the reflection responses or the original electromagnetic fields with the goal of extracting the conductivity model of the subsurface. For the inversion, a genetic algorithm and a nonlinear conjugate-gradient algorithm were used. Our results show that an inversion of the reflection responses produces worse estimates of the vertical conductivity but superior estimates of the horizontal conductivity (especially for the reservoir) with respect to the original electromagnetic fields.

Download Full-text

Data-driven Learning Algorithm of Neural Fuzzy Based Hammerstein-Wiener System

Journal of Sensors ◽

10.1155/2021/8920329 ◽

2021 ◽

Vol 2021 ◽

pp. 1-11

Author(s):

Feng Li ◽

Yinsheng Luo ◽

Naibao He ◽

Ya Gu ◽

Qingfeng Cao

Keyword(s):

Finite Impulse Response ◽

Learning Algorithm ◽

Gradient Algorithm ◽

Data Driven ◽

Noise Model ◽

Learning Approach ◽

Learning Technology ◽

Wiener System ◽

Neural Fuzzy ◽

Linear Block

A novel data-driven learning approach of nonlinear system represented by neural fuzzy Hammerstein-Wiener model is presented. The Hammerstein-Wiener system has two static nonlinear blocks represented by two independent neural fuzzy models surrounding a dynamic linear block described by finite impulse response model. The multisignal theory is designed for employing Hammerstein-Wiener system to separate parameter learning issues. To begin with, the output nonlinearity parameters are learned utilizing separable signal with different amplitudes. Furthermore, correlation analysis method is implemented for estimating linear block parameters using separable signal inputs and outputs; thereby, the interference of process noise is effectively handled. Finally, multi-innovation learning technology is introduced to improve system learning accuracy, and then, multi-innovation extended stochastic gradient algorithm is obtained for optimizing input nonlinearity and noise model using multi-innovation technique and gradient search method. The simulation results display that presented data-driven learning approach has the availability of learning Hammerstein-Wiener system.

Download Full-text