A robust and novel regression based fuzzy time series algorithm for prediction of rice yield

Time-Series archetypes have been administered for prediction in variegated and disparate domains like academic enrollments, weather forecasting, rice production etc., wherein an observation is contrived or excogitated based upon its past observation. The cogitation of fuzzy time series was first unveiled by Song & Chissom. This paper is the result of motivation from our previously published research work, wherein we delineate a regression based fuzzy time series approach, which deals with qualm, obfuscation, veracity, and spuriousness the various facets of the fuzzy contexture and furnishes a better and a more veracious result than the methods that are de facto and indubitable. Frequency based partitioning has been used as the partition of discourse and actual production as the universe of discourse. Subsequently, Fuzzy Logical Relationships of varying degrees have been executed to effectuate the fuzzification process. In addition to this, A Regression Analysis Model has been enforced to accomplish the defuzzification operation. To expound the process of forecasting, we have used the historical data of rice yield of University of Agriculture and Technology, India.