Measuringweather prediction accuracy using sugeno based Adaptive Neuro Fuzzy Inference system, grid partitioniong and guassmf

Today, Accurate weather prediction has been one of the major challenging environmental problems of the modern world. Estimates of temperature values are not only is an important factor in the agricultural decision making process, but also needed for environmental and technical applications i.e. assessment of natural disaster or crop growth forecasting. Several data mining techniques in collaboration with Artificial intelligence and statistical techniques are in use for this forecasting task. Due to the fuzzy nature of weather data, this paper solves weather event puzzle for the known industrial city of Pakistan, Sialkot, by implementing a fuzzy rule based system using Sugeno Fuzzy Inference. Two separate experimental settings have been used in this paper. To develop a fuzzy inference system, the first experimental data set consisting of 2100 instances with 14 inputs and 5 weather events. The second data set also consisting of 2100 instances but with of 6 input parameters. Finally comparative analysis of both experiments is done. Experimental results indicated that the accuracy of the both experiments demonstrate an increasing shift with an increase in the membership functions.