Using Machine Learning to Identify Factors of Importance in Aquaculture Management in Malaysia

The aquaculture business has expanded over the years in Malaysia. It provides a source of income and self-employment for many small-time players besides providing the protein requirement and a contributor to food security. However, the sustenance of this important sector is at stake as many small-timers are not making sufficient profits, citing increasing costs and dumbing from neighbouring countries. Hence, the focus of the study was to investigate various factors that influence this sector and identify important factors to help make this sector a viable and profitable venture in Malaysia. The various factors for the investigation were identified from the literature as influencers, and through the use of machine learning techniques on logit regression, significant factors were identified in the local context. The responses for the factors were adduced from a questionnaire survey directed at 268 farm operators and owners of aquaculture. The data collected from this survey form the primary input for the analysis. The important factors identified as drivers of Profitability are the Provision of Extension Services, Climate Change, Innovative Technologies, Farming Practices, Institutional Influences, Environmental, Learning and Development, and Economics. However, Societal, Supply Chain, Risk Management Culture, and Feed are not significant drivers in the local context. This study is limited in scope as it involves only the farm operators and owners and no other stakeholders and uses a single model in the machine learning process namely logit regression. However, findings are not diminished in the sense that the factors identified can be a source of input for policymakers in the local context for any future blueprint for this sector.