Application of the Neuro-Fuzzy Model in Security Feeling Analysis of Worn Tissues Residents (Case Study: Tehran Metropolis 12 District)

Security is a path to stability and an integral part of sustainability. Individuals depend on security to realize all their potential. Therefore, the purpose of the present research is security feeling analysis of worn tissues residents in Tehran metropolis 12 district based on social, economic and physical indicators.

The method of this research is descriptive-analytical and it uses regression and neuro-fuzzy method.The population consisted of 76,628 residential units in the 12th district. Using Cochran formula with 5% error and 95% confidence level, 383 units were selected as sample size.

Security feeling is predictive component strongest with beta coefficient (0.623). They are followed by lighting, passages, patterns of activity and uses, quality of housing, space form, citizens' identity, population density, social surveillance, employment and housing components. Their beta coefficients are 0/563, 0/509, 0/501, 0/497, 0/493, 0/491, 0/423, 0/319, 0/302 and 0/218 respectively.

There is a direct relationship between physical, economic and social indicators with feeling of security in Tehran metropolis 12 district.Based on the neuro-fuzzy model, data physical indices such as insecurity of area passages and high level of infiltration of motorcyclists to pedestrians have the lowest error rates in test and training. Social indicators such as overcrowding in some parts of District 12 and the presence of backyard space have the lowest error rates in test and training data. Economic indicators such as rising unemployment in the region and rising living costs in the have the lowest error rates of test and training data in Tehran metropolis 12 district.