Abstract: Due to its complexity and durability, precipitation prediction has recently gained the highest research relevance. Among other applications, such as flood forecasting and pollutant concentration monitoring. Existing model uses a complex statistical model that is often too expensive for both calculations and budgets. It does not apply to downstream applications. Therefore, an approach using machine learning algorithms. It is being studied in combination with time series data as an alternative to overcome these shortcomings. To this end, this study presents a comparative analysis based on a simplified precipitation estimation model. Efficient traditional machine learning algorithms and deep learning architectures for this Downstream application. This paper presents a time-series method called Neuralprophet for predictions of Maharashtra’s ten most popular cities. This method provides an estimate of rainfall using different atmospheric parameters like average temperature and cloud cover to predict the rainfall. The main advantage of this model is that this model estimates the rainfall based on the previous correlation between the different atmospheric parameters. Thus, an estimated value of what the rainfall could be at a given period and place can be found easily. Introducing Neural Prophet, the successor to Facebook Prophet, which sets the industry standard for a explainable, scalable, and easy-to-use prediction framework. NeuralProphet is a hybrid prediction framework based on PyTorch and trained using standard deep learning techniques, so developers can easily extend the framework. Local context is introduced in autoregressive and covariate modules that can be configured as classical linear regression or neural networks. It includes traditional statistical and neural network models for time series modeling used for forecasting and anomaly detection. This model produces high-quality predictions of time series data showing multiple seasonality’s with linear or non-linear growth. Use this model to predict future temperatures in Maharashtra's most popular cities using historical temperature data from the same location.
|
DOI:
10.17148/IJARCCE.2022.11573
