Forecasting Bangladeshi Garlic Price Using DL and ML Hybrid Approach

Market uncertainty is a constant matter in Bangladesh.Therefore, the prices of our regular substances fluctuate each so frequently.It significantly influences the constituents we consume daily.Almost every meal contains garlic in Bangladesh.People require to keep a record of the price of the materials they use every day but arranging it manually is a strenuous task.Machine Learning (ML) and Deep Learning (DL) permit us to accomplish such a task with much consequence.In this work, we established our dataset from the particulars obtainable from the Ministry of Agriculture, Bangladesh.For predicting the price of garlic, we employed four main architectures based on the DL approach.The first model is the Deep Neural Network (DNN) model, and the second and third models are LSTM-based.Finally, the fourth model is LSTM+ML combined hybrid architecture.By proper analysis, we estimated the best-performing algorithm for the prediction of garlic price.Different matrix results were delivered by the algorithms and performance difference is not so much among all the algorithms.The highest r2 score is achieved by model 3 with a score of 95.83%.