MEAN SQUARED ERROR APPLIED IN BACK PROPAGATION FOR NON LINEAR RAINFALL PREDICTION
In one sentence. Multiple linear regression feeds a backpropagation neural network to forecast low, average and heavy rainfall at Bengaluru from five years of station data.
Abstract
Analyzing global weather forecasts is challenging and expensive. Machine learning algorithms analyze trends in weather data by adopting regression model and neural network model. Our proposed model is based on two methodologies Multiple Linear Regression (MLR) and Artificial Neural Network (ANN) to predict rainfall. MLR determines most significant parameters of rainfall data for ANN. Mean Squared Error (MSE) generated at ANN model was back propagated to get more accurate results. The model was tested on five year (2013 to 2017) meteorological data of Bengaluru station. ANN with Back Propagation Neural Network (BPNN) was applied to forecast low rainfall, average rainfall and heavy rainfall. Performance of the model was measured by R and MSE value. Experimental result shows that back propagation of MSE and data normalization yields good results.
Labels
- Topic
- Local & station forecasting
- Type
- New model or method
- Method
- Classical ML (trees, GP, linear, kernels)
- Time scale
- 6 h–3 days
- Domain
- Local / station
- Data
- Station / in-situ
- Status
- Research
- Identifiers
- DOI:10.5281/zenodo.14911970 OpenAlex:W3043540852
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.