A Novel Approach for Weather Prognosis Using Naive Bayes Modeling Technique over Srinagar, J&K, India

Weather prognosis has become a very important in terms of global economy of the country, agriculture and industries. The accurate prediction of precipitation can sense the heavy rainfall and can provide warnings and information regarding the disasters that could happen in future. Various techniques were proposed and implemented to forecast precipitation, but did not attain much accuracy due to changing weather conditions. In this research work we have tried to implemented Naive Bayes algorithm to build rainfall forecast model which will forecast the precipitation with significant accuracy. Historical weather data set of Srinagar, J&K, India, is gathered from November 2015 to November 2016 from http:///www.wundergrounds.com website. From nine (9) available set of attributes; Temperature, Humidity, sea level pressure, wind speed and Events attributes are considered for this work to achieve better results. Experimental outcomes of various performance measures illustrate that Naive Bayes approach to forecast precipitation has appreciable accurateness and credibility.