Functional gradient descent for n-tuple regression

Katopodis, Rafael F.; Lima, Priscila M. V.; França, Felipe M. G.

doi:10.1016/j.neucom.2022.05.114

Search citation statements

Order By: Relevance

Paper Sections

Select...

State and Encoder1

Introduction1

Citation Types

Supporting

Mentioning

Contrasting

Year Published

2023

2024

Publication Types

Select...

Other1

Article1

Relationship

Self Cite0

Independent2

Authors

Journals

Cited by 2 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…N-tuple Regressors have the advantage of combining low usage of computational resources in training and prediction, with the high capability of generalization with a low number of hyperparameters. In this context, is important to include a forgetting mechanism in order to deal with the changes in the Q distribution [12]. Policy The policy uses the Q values predicted for each action to choose how to act.…”

Section: State and Encodermentioning

confidence: 99%

See 1 more Smart Citation

Sun Tracking using a Weightless Q-Learning Neural Network

Souza,

Lima,

França

2023

ESANN 2023 Proceesdings

View full text Add to dashboard Cite

Photovoltaic(PV) systems are one of the leading technologies to address climate change. Tracking systems improve energy generation by moving the surface to follow the sun's position however, these methods do not ensure optimal results in cloudy environments. This article proposes a closed-loop control algorithm for tracking based on reinforcement learning and weightless neural networks, compared to an astrological model. The method was applied in a single PV array on a single-axis tracking system, simulated with PVLib. Results showed that the architecture could improve results in cloudy environments but not in a clear-sky situation, as expected for a first approach.

show abstract

Section: State and Encodermentioning

confidence: 99%

“…The prediction process is the average of the outputs of repeating patterns in the input. Katopodis et al [12] proposed the inclusion of a forgetting mechanism using a discounting rate, which improves the performance of the model in sequential decision contexts, such as reinforcement learning.…”

Section: Introductionmentioning

confidence: 99%