Students apply artificial intelligence to create coffee price forecasting model

A group of RMIT Vietnam students developed a model that can forecast Robusta coffee prices by leveraging historical data on coffee prices, petrol prices, temperature and rainfall.

Vietnam is the world's second largest coffee exporter and accounts for more than half of the global Robusta supply. Coffee production in the 2022/23 crop year is expected to reach 29.75 million bags, of which Robusta accounts for more than 95%. However, prices of agricultural products in general and coffee beans in particular are often unstable and can fluctuate sharply during bumper harvests, significantly affecting farmers' incomes and causing damage to the economy.

A group of final-year Bachelor of Information Technology students from the Faculty of Science, Engineering and Technology, including Nguyen Hai Minh Trang, Doan Chanh Thong, Le Ngoc Nguyen Thuan, Nguyen Phuong Nam and Lam Tin Dieu, along with their instructors, trained and evaluated six machine learning (ML) models to predict coffee prices. The models can help Vietnamese farmers make appropriate crop decisions and planning, optimizing profits and minimizing losses.

The RF model gives the best results. Photo: NVCC — The RF model gives the best results. Photo: *NVCC*

Nguyen Hai Minh Trang, a research member, said the team developed six ML models, namely LSTM, GRU, ARIMA, SARIMA, SVM and RF, based on the history of coffee prices, gasoline, temperature and rainfall, to predict Robusta coffee prices in Lam Dong province. It was found that the RF model, which uses the entire dataset, was the most effective.

The reason is that RF can incorporate richer datasets and handle nonlinear relationships. In addition, fuel price is shown to be a significant predictor and outperforms all other tested features combined.

According to the research team, the model has the potential to be further improved by studying and adding the impact of crop yields, market trends and geopolitical events on agricultural prices.

Research team members. Photo: NVCC — Research team members. Photo: *NVCC*

The research results were presented at the 8th IEEE/ACIS International Conference on Big Data, Cloud Computing and Data Science Techniques (BCD 2023) - together with researchers, scientists, engineers and experts held in December in Ho Chi Minh City. Here, experts gave suggestions to improve the accuracy and applicability of the model's predictions. "We plan to delve deeper into advanced techniques and emerging methods in this field to further strengthen the research results that the team has done," Thong said.

Hai Minh

Source link