AlbeeBrew/CST-Masters-Dissertation-Crop-Yield-Prediction

Using machine learning methodologies to predict crop yield for some Southern and Northern Ghana. # CST-Masters-Dissertation-Crop-Yield-Prediction Using machine learning methodologies to predict crop yield for some Southern and Northern Ghana. Climate change has been a global issue for decades. Climate change affects crop quality and contributes to food insecurity. To help mitigate the problem of climate change related food insecurity in Ghana, machine learning algorithm, which has proven in many areas to be effective in making predictions was proposed in this study. Six food crops - rice, maize, cassava, yams, tomatoes, and plantain, were used in this study. The algorithm used climatic, weather, agricultural yields, and chemical data to forecast the production of these six food crops. The model was built using decision tree, logistic regression, and k-nearest neighbor algorithms. The three machine learning models generated encouraging results. Hyper-parameter adjustment was used to avoid overfitting. The decision tree model had an of 95.3%, whereas the K-Nearest Neighbor and logistic regression had values of 93.15% and 89.78% respectively. The correlation between the projected and the expected outcome is supported by the decision tree model while the K-Nearest Neighbor model was correlated with the expected data.