A long short term memory based weather forecast prediction using novel linear regression with neural network model over support vector machine.

Using the suggested Novel Linear Regression Algorithm with Neural Network Model, the purpose of the study is to make a Long-short Term Memory (LSTM)-based weather forecast prediction. Both the Materials and the Methods: Sample groups that are taken into consideration for the project can be divided into two categories: one for the Linear Regression (LR) and another for the Support Vector Machine (SVM). Both of these sample groups are tested with a G-power of 0.80 to determine the sample size and for a T-test analysis. More than 4188 records were utilised in the testing of the datasets in order to get the forecast. The conclusion reached by the researchers was that the Linear Regression Algorithm, which has an average accuracy of 94% and appears to be superior than the SVM, which delivers around 91%, as a result of their research. The significance is somewhere around 0.329, and since the p-value is more than 0.05, we may conclude that there is no statistically significant difference between the research group. The LR algorithm appears to be superior to the SVM method in terms of its ability to forecast rainfall. [ABSTRACT FROM AUTHOR]