Deep learning in data science

Abstract

Up until early 2000’s climate predictions were made mainly using statistical methods. This prediction wasn’t always entirely accurate. With the introduction of deep learning in climate prediction, the prediction accuracy has improved dramatically. The sensors in the weather stations give massive amount of unstructured data. Due to the humungous amounts of sensors and data from it, it’s almost impossible to compute all the necessary weather information in time. AI and deep learning help to overcome this problem using different models which can swiftly and accurately make this job simple. Accurate climate prediction is very important to predict is very important to predict any natural calamities or unexpected change in weather. This report highlights few of the deep learning models which can be used for climate prediction by scientists. This paper only takes scratches the surface of the capabilities of AI in climate change. More advancements in this field would lead to better simulations of the weather conditions which can then be useful to predict the extreme weather conditions accurately. Few of the authors have used unique models in their prediction of various temperature, rainfall, pollution levels etc. which have helped them to find the discrepancies in the climate if any.