What is it about?

Countries depending on agricultural crops have significant impacts on the world's challenging issues such as world hunger and malnutrition, poverty, insufficient food supply and even climatic changes. Governments who realize the importance of agriculture towards their respective countries are able to seize the opportunity to involve more people in the agricultural industry. However, there have been several factors that put off people from involving in agricultural activities as this group of people are struggling to make a profit out of their hard work. The factors include lack of experiences amongst farmers in doing farming activities, as well as instability of price flow impact from environmental and economic factors. This paper proposes a software application with price forecasting features to assist farmers in sufficing their agriculture market knowledge and maximize their profit in doing farming business. This paper investigates different trending machine learning algorithms, such as ARIMA, LSTM, SVR, Prophet and XGBoost and then undergoes an experiment to identify the most optimal algorithm as a model for the software. From the MSE result, LSTM is discovered to be the most accurate and efficient in handling increasing amounts of complex data. Besides forecasting features, the software is also utilized with weather information, agriculture news, and also cooperating with third parties' suppliers to create a platform for farmers' trading.

Featured Image

Why is it important?

Countries depending on agricultural crops have significant impacts on the world's challenging issues such as world hunger and malnutrition, poverty, insufficient food supply and even climatic changes. Governments who realize the importance of agriculture towards their respective countries are able to seize the opportunity to involve more people in the agricultural industry. However, there have been several factors that put off people from involving in agricultural activities as this group of people are struggling to make a profit out of their hard work. The factors include lack of experiences amongst farmers in doing farming activities, as well as instability of price flow impact from environmental and economic factors.

Perspectives

This paper proposes a software application with price forecasting features to assist farmers in sufficing their agriculture market knowledge and maximize their profit in doing farming business. This paper investigates different trending machine learning algorithms, such as ARIMA, LSTM, SVR, Prophet and XGBoost and then undergoes an experiment to identify the most optimal algorithm as a model for the software. From the MSE result, LSTM is discovered to be the most accurate and efficient in handling increasing amounts of complex data. Besides forecasting features, the software is also utilized with weather information, agriculture news, and also cooperating with third parties' suppliers to create a platform for farmers' trading.

Dr zhiyuan chen
The University of Nottingham Malaysia

Read the Original

This page is a summary of: Long Short-Term Memory Model Based Agriculture Commodity Price Prediction Application, August 2020, ACM (Association for Computing Machinery),
DOI: 10.1145/3417473.3417481.
You can read the full text:

Read

Resources

Contributors

The following have contributed to this page