Weather Forecasting System

Ashley Mohanani, Rucha Kalode, Aishwarya Dhumane, Aman Zade, Shweta Hiwanj, C B Tripathi

Zenodo (CERN European Organization for Nuclear Research) · 21 March 2020

In one sentence. Compares RNN, LSTM, GRU, linear regression and random forest for station relative-humidity prediction from date and time inputs, ranking which gives best accuracy under limited resources.

Abstract

The Meteorological Department is an agency that takes readings of meteorological observations, weather forecasting and seismology. Universally, there are three types of weather prediction techniques, namely, synoptic weather prediction, numerical weather prediction, statistical weather prediction. The presented paper focuses on statistical weather prediction using the concept of artificial neural networks and regression algorithms. Recurrent Neural Network (RNN), Long Short Term Memory (LSTM), Gated Recurrent Unit (GRU), Linear Regression and Random Forest Regression, are the algorithms that are used in the study, where the first three are categorized as ANN algorithms and remaining two are regression algorithms that do not include neural networks. A comparison is shown between all the five algorithms and their predicted value of relative humidity, with date and time as input values. This paper is used as a study of which algorithm provides best accuracy with the limited time and resource constraints.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.