FORCASTING NATURAL GAS PRICE PREDICTION USING MACHINE LEARNING

FORCASTING NATURAL GAS PRICE PREDICTION USING MACHINE LEARNING

 1Harish B G,2Chetan kumar G S,3Raghavendrareddy  Radder, 4Manoj K,5Shivarajgouda NB, 6Soniya R S

 1Assistant Professor, Department of Master of Computer Application, UBDTCE, Davangere

2,3,4,5,6 Student, Department of MCA, UBDTCE, Davangere

________________________________________________________________________________________________

Abstract :   In this paper, we will use statistical modeling techniques to forecast natural gas prices in US markets. We will investigate the application and contrast of the decision-tree and random forest linear regression models. Using a total of 5800 records, we used a dataset of natural gas prices in the US expressed as US dollars per Btu from 1998 to 2020. Random forest conducts regression using ensemble learning techniques since it is a supervised learning algorithm. Random forest will forecast the price by utilizing the current prices and comparing them, which will be useful for the user to determine the risk in bidding prices. The results are indicated with up to 97% accuracy. The stage has been prepared for the potential new opportunities that these technologies may present.

Keywords   : Random Forest, Decision Tree, Prediction, Natural Gas.