Volume 20 No 12 (2022)
 Download PDF
BIPARTITE QUADRATIC FISHER SCORE AND TRIPLET LOSS BASED DEEP LEARNING FOR SOIL QUALITY PREDICTION
Balaji.G, Dr.P.Vijayakumar
Abstract
The vital center of attention to the management of soil to improve crop productiveness was a conservation as well as refinement of vigorous soil. Some of reasons like, population pressures, terrestrial disadvantages as well as conventional soil mechanisms are charge of to worsening or declining in soil fertility as far as developing countries like India. Hence, a considerable improvement within crop is arrived at applying appropriate crop fitness technique. On other hand, an increase in productivity is said to be arrived at via efficient soil resource management and remedial assessments. Timely identification issues associated using soil management for ensuring productivity. Over the past few years, classification issues were efficiently employed as ML and Deep Learning (DL) techniques. Therefore, identifying the accurate and precise method to predict the soil quality is still the research of interest. Bipartite Quadratic Fisher Score as well as Triplet Loss-based Deep Learning (BQFS-TLDL) is proposed for predicting the soil quality in an accurate and precise manner. First, with the soil moisture prediction big dataset provided as input, significant and precise features are selected by means of Bipartite Quadratic Mutual Information and Fisher Score model. Next, with the significant features selected, Triplet Loss-based Deep Learning is applied for predicting soil quality. Simulation of deep learning methods, BQFS-TLDL has better accuracy as well as efficiency with minimum error rate
Keywords
Machine Learning, Bipartite, Quadratic, Mutual Information, Fisher Score, Triplet Loss, Deep Learning, Soil Quality Prediction
Copyright
Copyright © Neuroquantology

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Articles published in the Neuroquantology are available under Creative Commons Attribution Non-Commercial No Derivatives Licence (CC BY-NC-ND 4.0). Authors retain copyright in their work and grant IJECSE right of first publication under CC BY-NC-ND 4.0. Users have the right to read, download, copy, distribute, print, search, or link to the full texts of articles in this journal, and to use them for any other lawful purpose.