Loading...
Theses
Browse
1 results
Search Results
ThesisItem Open Access A statistical study of temperature forecasting data(G.B. Pant University of Agriculture and Technology, Pantnagar - 263145 (Uttarakhand), 2020-12) Abhishek, S.S.; Bhardwaj, S.B.In our daily life, we often use some forecasting techniques to predict weather, temperature, economy, etc. Based on these forecasting results, we can prevent damages to occur or get benefits from the forecasting activities. In fact, an event in the real-world can be affected by many factors. The more the facts we consider, the higher the forecasting accuracy rate. Moreover, the length of each interval in the universe of discourse also affects the forecasting results. The present research work entitled ‘A statistical study of temperature forecasting data’ has been undertaken to compute errors empirically by considering the actual temperature along with corresponding 36 months forecasted observations. Moreover, an attempt has been made to measure the forecast accuracy with actual maximum and minimum temperatures with corresponding hypothetical temperatures of last 36 months using Theil’s U-statistic. A low value of Theil’s statistic is revealing that there is a greater forecasting accuracy for both the actual minimum and maximum temperatures and the corresponding hypothetical forecasted observations. In the final segment of this study, identification of the best fitted model using exponential smoothing method through coefficient of determination (R2), mean absolute error (MAE) and mean absolute percent error (MAPE) has also been done. The model Holtwinters’ additive model was found to be best fitted model as it has exhibits highest R2 and lowest MAE and MAPE values. The study was carried out for Pantnagar, U. S Nagar, Uttarakhand using time series data from 2017 to 2019.
