Affiliation:
1. Kamrup Polytechnic, India
Abstract
Weather forecasting is a scientific method that involves the prediction of atmospheric conditions at a specific geographic location. The increased volatility over the last decade is owing to an enormous rise in water used for irrigation application throughout agricultural area, much of which evaporates, necessitating accurate forecasting in order to take essential safeguards. In this chapter, an attempt is made to predict the average temperature and maximum temperature through machine learning models. The daily temperature data from 1970 to 2022 were collected from the National Centre of Environmental Information (NCEI). The ARIMA model is used to predict the weather data sets of Tezpur, Assam. Previous conventional models are insufficient to predict forecasting precisely. So statistical models and auto-regressive models are programmed and compared.
Reference20 articles.
1. A comparative study of series hybrid approaches to model and predict the vehicle operating states
2. Weather Prediction using Multi Linear Regression algorithm. Material Science and Engineering.;N.Anusha;IOP Conference Series,2019
3. BoxG. E.JenkinsG. M.ReinselG. C. (2011). Time series analysis:Forecasting and Control. John Wiley and Sons.
4. Chai, T., & Dexler, R. R. (2014). Root mean Square error or Mean Absolute Error-Arguements avoiding RMSE in the literature. Geoscience model devlopment, 7(3), 1247-1250.
5. Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014). Empirical evaluation of Gated Recurrent neural network on Sequence Modelling. arxiv.