Attribute Selection for Earths Climate Prediction

Abstract: The strong association between monthly average rainfall and monthly mean temperature over the 30-year period from 1976 to 2006 for Vindhya Region is analyzed. This study shows climatological characteristics and fluctuation of climate with rainfall and found temperature as a most significant attribute. Both local and field significances have been tested by using CFS subset evaluator. Greedy algorithm applies on the normalized data sets of selected three solar cycles. From the analysis we found that the merit of best subset are 0.394 (SC-21), 0.274 (SC-22) and 0.348 (SC-23) for selected solar cycles. We have observed that for all the solar cycles temperature is the most significant attribute among the selected solar activity parameters.

Keywords: Greedy algorithm, temperature, CFS evaluator, wind velocity

Edition: Volume 5 Issue 8, August 2016,

Pages: 1876 - 1878

How to Download this Article?

Type Your Valid Email Address below to Receive the Article PDF Link

Verification Code will appear in 2 Seconds ... Wait

Attribute Selection for Earths Climate Prediction

Similar Articles with Keyword 'temperature'

Recent Trends in Surface Air Temperature over Al-Taif, Saudi Arabia

Case Study of Secondary Radioclimatic Variables on Radio Signals at the Troposhere in Nsukka, Nigeria