Enhancing the Implementation and Reliability of Nanomaterial Detectors through Deep Learning Optimization

Authors M.A. Jawale1, P. William1 , N.K. Darwante2, V. Verma3, S.S. Ingle4 , D. Roy5
Affiliations

1Department of Information Technology, Sanjivani College Engineering, Kopargaon, SPPU, Pune, India

2Department of Electronics & Computer Engineering, Sanjivani College of Engineering, Kopargaon, SPPU,Pune, India

3Department of Computer Science and Engineering, Bhilai Institute of Technology, Raipur, Chhattisgarh, India

4Department of Mechanical Engineering, Sanjivani College Engineering, Kopargaon, SPPU, Pune, India

5Hyderabad Institute of Technology and Management, Gowdavelli Village, Medchal, Hyderabad, India

Е-mail william160891@gmail.com
Issue Volume 16, Year 2024, Number 5
Dates Received 10 June 2024; revised manuscript received 17 October 2024; published online 30 October 2024
Citation M.A. Jawale, P. William, et al., J. Nano- Electron. Phys. 16 No 5, 05002 (2024)
DOI https://doi.org/10.21272/jnep.16(5).05002
PACS Number(s) 07.05.Mh
Keywords Deep Learning (DL), Nanomaterials (NMs), Nanotechnology (6) , Fine-tuned genetic algorithm-based dynamic deep neural network (FTGA-DDNN).
Annotation

Improving the reliability and implementation are critical in real-world applications, and the inherent unpredictability of non-materials renders it complicated to integrate Nanomaterial (NMs) detectors into these environments. Reliable presumptions can be constructed based on the data produced by such sensors using Deep Learning (DL), which is a potent method. In this study, we proposed a novel method called Fine-Tuned Genetic Algorithm Based Dynamic Deep Neural Network (FTGA-DDNN) which is computationally costly to train, yet it yields the most efficient result when evaluated the internet, maintaining a reasonable level of reliability. This can be beneficial in dynamically changing environments where the algorithm needs to explore new possibilities while exploiting known solutions. Through DL optimization, the goal of improving the implementation and dependability of nano-material detectors is to increase their adaptability and efficacy in a variety of situations. We present a comparative analysis of the results obtained from our proposed technique against other existing methods. Our findings indicate superior performance in average error, average absolute error, and semi-log testing time, showcasing the efficacy of the FTGA-DDNN approach. In summary, this allows us to forecast and predict the filter function later on, improving the DL algorithms' accuracy and the filters' usefulness over extended periods.

List of References