Application of Machine Learning to Investigate the Impact of Climatic Variables on Marine Fish Landings
The fisheries industry of Malaysia is known as the strategic sector that can help the country raise domestic food production and supply. This research proposed machine learning (ML) based prediction of marine fish landings to project fish supply and compare those projections with the observed data....
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Published: |
Springer
2023
|
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Tenaga Nasional |
| Summary: | The fisheries industry of Malaysia is known as the strategic sector that can help the country raise domestic food production and supply. This research proposed machine learning (ML) based prediction of marine fish landings to project fish supply and compare those projections with the observed data. Three ML models, i.e., linear regression (LR), decision tree (DT), and random forest (RF) regression, are applied to the dataset that contains 18�years of climatic variables and the marine fish landings (tonnes) information of 5 major states of Malaysia. The results suggest that the developed LR model shows an R2 value of 0.60 and 0.64 in the validation and testing phases. The DT and RF model indicates a significant improvement as the R2 values are 0.88 and 0.89 in the validation data and 0.89 and 0.86 in the testing data. Finally, we calculated the Nash�Sutcliffe efficiency (NSE) values, and the results indicated that RF based ML model has the highest NSE value of 0.86, which turns out to be the best fit for prediction. The developed ML models have utilized for the first time to predict the marine fish landing using environmental inputs collected from 5 different states of Malaysia. � 2022, The Author(s), under exclusive licence to The National Academy of Sciences, India. |
|---|