sensitivity analysis of economic variables using neuro- fuzzy approach

Sensitivity analysis (sa) is a vital task for decision making in economic management. in this paper, a novel fuzzy sensitivity analyzer (fsa) is proposed to analyze the sensitivity of economic variables. the proposed fsa algorithm consists of an adaptive neuro-fuzzy inference system (anfis) that is adjusted for forecasting economic time series. based on the output of anfis, fsa can determine the importance degree of parameters. in the numerical studies, the proposed method is applied for the sensitivity analysis of oil and gold time series. according to the results, fsa indicates that oil price is highly dependent upon the inflation rate, dollar index and market index while opec production level and gold price have less impact. furthermore, in the gold price modeling, the highest sensitivity is obtained from silver price while demand for gold is more a function of market index and inflation rate. the proposed method can be used in many sa applications.