LC oscillator frequency prediction using machine learning linear regression algorithm
Uloženo v:
| Název: | LC oscillator frequency prediction using machine learning linear regression algorithm |
|---|---|
| Autoři: | Mandar Jatkar, Vasudeva G, Tripti R. Kulkarni, Roopa R. Kulkarni, Aneesh Pandurangi |
| Zdroj: | Discover Electronics, Vol 2, Iss 1, Pp 1-11 (2025) |
| Informace o vydavateli: | Springer, 2025. |
| Rok vydání: | 2025 |
| Sbírka: | LCC:Electrical engineering. Electronics. Nuclear engineering LCC:Mathematics |
| Témata: | Linear regression, Frequency prediction, LC circuits, Machine learning, Electrical circuits, Inductance, Electrical engineering. Electronics. Nuclear engineering, TK1-9971, Mathematics, QA1-939 |
| Popis: | Abstract The research demonstrates a new method for predicting LC circuit frequencies through linear regression techniques applied to standard resonance formula-derived log-transformed L and C values. The modeling technique produces outstanding precision because it delivers a root mean squared error measurement of 1.51 × 10⁻¹². The observed relationship reveals that frequencies affect component measurements by -0.5 while using − 1.8379 as the intercept and matching the predictive resonance model. The performance visuals of the proposed model show minimal error in prediction across different LC combinations along with smooth and reliable predictive outputs and negligible bias according to the residual histogram. We examine different regression methods to determine the resonant frequency of LC circuits by transforming the input features into log values. It was found that Linear Regression reaches close to zero RMSE, but SVR and Random Forest show higher errors. The study proves that log-transformed linear regression works as a basic but efficient method for accurate frequency estimation in resonant LC circuits. |
| Druh dokumentu: | article |
| Popis souboru: | electronic resource |
| Jazyk: | English |
| ISSN: | 2948-1600 |
| Relation: | https://doaj.org/toc/2948-1600 |
| DOI: | 10.1007/s44291-025-00092-9 |
| Přístupová URL adresa: | https://doaj.org/article/ab3e3e6279344769ae6d8fe0a312c036 |
| Přístupové číslo: | edsdoj.b3e3e6279344769ae6d8fe0a312c036 |
| Databáze: | Directory of Open Access Journals |
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstrakt: | Abstract The research demonstrates a new method for predicting LC circuit frequencies through linear regression techniques applied to standard resonance formula-derived log-transformed L and C values. The modeling technique produces outstanding precision because it delivers a root mean squared error measurement of 1.51 × 10⁻¹². The observed relationship reveals that frequencies affect component measurements by -0.5 while using − 1.8379 as the intercept and matching the predictive resonance model. The performance visuals of the proposed model show minimal error in prediction across different LC combinations along with smooth and reliable predictive outputs and negligible bias according to the residual histogram. We examine different regression methods to determine the resonant frequency of LC circuits by transforming the input features into log values. It was found that Linear Regression reaches close to zero RMSE, but SVR and Random Forest show higher errors. The study proves that log-transformed linear regression works as a basic but efficient method for accurate frequency estimation in resonant LC circuits. |
|---|---|
| ISSN: | 29481600 |
| DOI: | 10.1007/s44291-025-00092-9 |