Recovering historical climate records using artificial neural networks in GPU

Abstract

This article presents a parallel implementation of Artificial Neural Networks over Graphic Processing Units, and its application for recovering his-torical climate records from the Digi-Clima project. Several strategies are intro-duced to handle large volumes of historical pluviometer records, and the paral-lel deployment is described. The experimental evaluation demonstrates that the proposed approach is useful for recovering the climate information, achieving classification rates up to 76% for a set of real images from the Digi-Clima pro-ject. The parallel algorithm allows reducing the execution times, with an accel-eration factor of up to 2.15×.