Related Experiment Video
Updated: Sep 18, 2025
05:30
Large Scale Energy Efficient Sensor Network Routing Using a Quantum Processor Unit
Published on: September 8, 2023
669
Fertilizer prediction using serial exponential newton meta-heuristic algorithm-based convolutional neural network in
1Computer Science and Engineering, Mar Ephraem College of Engineering and Technology, Elavuvilai, Marthandam, KanniyakumariDist, Tamil Nadu 629171, India; Computer Science and Engineering, Mar Ephraem College of Engineering and Technology, Elavuvilai, Marthandam, Kanniyakumari Dist, Tamil Nadu, India.
Computational Biology and Chemistry
|June 26, 2025
Summary
This study introduces a Deep Learning model for fertilizer prediction in IoT Wireless Sensor Networks. The novel SExpNMA algorithm enhances 1D CNN performance, achieving high accuracy in predicting fertilizer needs.
Area of Science:
- Agricultural Technology
- Machine Learning
- Wireless Sensor Networks
Background:
- Optimizing fertilizer application is crucial for sustainable agriculture and crop yield.
- IoT-based Wireless Sensor Networks (WSN) generate vast amounts of data for agricultural monitoring.
- Existing methods for fertilizer prediction may lack efficiency and accuracy in complex WSN environments.
Purpose of the Study:
- To develop and evaluate a Deep Learning model for accurate fertilizer prediction using IoT-based WSN data.
- To introduce a novel meta-heuristic algorithm (SExpNMA) for optimizing routing and classifier parameters.
- To enhance the performance of fertilizer prediction models through advanced data processing and feature fusion techniques.
Main Methods:
- Serial Exponential Newton Meta-Heuristic Algorithm (SExpNMA) for Cluster Head selection and routing.
- Integration of Newton Meta-heuristic Algorithm (NMA) and Serial Exponential Weighted Moving Average (SEWMA).
- Data preprocessing including normalization, cleaning, and augmentation.
- Feature fusion using Bidirectional Long Short-Term Memory (BiLSTM) based on feature correlation.
- Fertilizer prediction using a 1D Convolutional Neural Network (1D CNN) trained by SExpNMA.
Main Results:
- The SExpNMA-based 1D CNN model demonstrated superior performance compared to conventional models.
- Achieved high prediction metrics: 94.2% specificity, 94.8% sensitivity, and 94.5% accuracy.
- Effective handling of data through normalization, cleaning, and augmentation improved model robustness.
Conclusions:
- The proposed Deep Learning approach, leveraging SExpNMA and 1D CNN, offers a highly accurate solution for fertilizer prediction in WSN.
- The SExpNMA algorithm effectively optimizes WSN routing and enhances classifier performance.
- This research contributes to precision agriculture by enabling more efficient and data-driven fertilizer management.
