Machine Learning Model Approaches for Price Prediction in Coffee Market using Linear Regression, XGB, and LSTM Techniques

Abstract

Investors and other business persons have a desire to know about the future market price because, if the investors know about the future price of a certain commodity or stock it will help them to make appropriate business decisions and they can also get profit out of their investment. There are many previous researches that has been done on stock market predictions but there is no related research that has been done on Ethiopia commodity exchange (ECX). Performing future price prediction with better accuracy and performing comparative analysis between the algorithms for two of Ethiopia commodity exchange (ECX) items which are Coffee and Sesame as the research key objectives. Three different types of prediction algorithms to predict the future price, such as Linear Regression (LR), Extreme Gradient Boosting (XGB), Long Short-Term Memory (LSTM) was utilized. There are limited researches worked on price prediction of ECX items specifically, the idea of the price prediction on different Stock markets like New York stock market Exchange and other commodity market items prediction in order to develop our research in ECX was presented. The study apart from predicting the future price, comparative analysis was implemented between the prediction algorithms that we used based on their performance. Two different datasets from ECX: coffee and sesame were used. The reason for the utilization of these datasets is, the commodity items are the largest export items in Ethiopia which makes them very important for Ethiopian economy, and the different datasets helps us to get the advantage of evaluating the algorithms with different number of datasets, since sesame dataset has 7205 instances and coffee dataset has 1540 instances and both of them has 11 attributes. We build an android application in order two implement our algorithms on mobile applications and see if it is possible to implement the prediction algorithms on mobile platforms and make it easy and accessible to users. We call this mobile application Ethiopia Coffee Prices Predictor (ECPP). This application will be used to display the prediction result of Ethiopia Coffee price for short period and it is designed in the way to be user friendly. The programming environment used to implement the prediction algorithms is python, java programming language to design our android application and we used PHP to implement the API, and finally we used MySQL database in order to store information’s online and make them accessible everywhere.

Mobile phone price prediction: A comparative study among four models

“…By employing the linear regression methodology to establish a more accurate model for mobile phone price prediction, so that people can better predict the price of mobile phones to choose more cost-effective mobile phones [5]. When employing machine learning techniques for predicting future mobile phone prices, linear regression and XGB Regressor methodologies can be utilized to construct models that enhance the accuracy of mobile phone price predictions [6]. In order to investigate the ranking of impact factors affecting the importance of mobile phone prices, investors can employ the variable influence method in random forest regressor.…”