Combined Neural Approach to Deterministic and Probabilistic Assets Cost Valuation

One of the specific particulars of assets cost valuation is multiple influencing factors nature which influences their cost. These determines reasonable uncertainty and probabilistic character of cost valuation in general since to get real effect of each influencing factor especially in cases of their interrelations is a complicated problem. In its turn this leads to low efficiency of classical methods of multi-factor regression analysis (MFRA) utilization for the purposes of assets cost valuation. Apart of multi-factor regression analysis one of the promising approach for the efficient solving of existing problem is the neural network application. Based on comparative analysis of results obtained with most efficient modifications of neural network approach in this article advantages which gives developed combined clusterneural model demonstrated. Conducted are comparative calculations of such combined approach, which gives more reliable results with probability related range of property cost as have been demonstrated in the article.