(1) parameter analysis method
The analysis process of this method is as follows: ① Select a set of data that can reflect the whole picture of regional investment environment, such as investment coefficient, investment multiplier, marginal consumption tendency, investment saturation, infrastructure adaptability, investment risk, effective demand rate, national consumption level, resource appreciation rate, optimized commodity rate, etc. ② Estimate the values of each parameter, and compare the estimated results with similar indicators in areas with good investment environment; ③ Analyze the similarities and differences of the comparison results, so as to determine the quality of the investment environment (similarity means a good investment environment). This method is comprehensive and objective, but it is difficult to determine "recognized areas with good investment environment" without clear explanation.
(2) Quasi-number analysis method
The factors that affect the investment environment are divided into investment environment incentive coefficient K, urban planning perfection coefficient P, tax and profit coefficient S, labor productivity coefficient L, regional basic coefficient B, exchange rate coefficient T, market coefficient M and management right coefficient F by the method of quasi-number analysis. Each type of factor can be further divided into several sub-factors. After weighting these sub-factors, the total score of this kind of factors can be obtained by summation. In order to reflect the organic relationship between one factor and other factors, comprehensive standards are usually used to measure the investment environment, so as to overcome the shortcomings of mechanical scoring method and get more comprehensive evaluation results.
(3) Analytic Hierarchy Process
Analytic Hierarchy Process (AHP) was put forward in 1980 by T.L.Saaty, a logistics scientist at the University of Pittsburgh. Based on the principle of system hierarchy in system theory, this method decomposes complex problems into several simple problems with orderly and organized levels, and is a multi-level weight analysis and decision-making method. On the other hand, it will analyze, compare, quantify and sort the simpler problems relative to the original problems, and then make a comprehensive evaluation and optimization step by step to deal with people's subjective judgments in a formal way. Analytic Hierarchy Process (AHP) is a multi-factor evaluation and optimization method, and a new method combining qualitative and quantitative analysis. There are five steps in the application of AHP: establishing hierarchical structure model, constructing judgment matrix, ranking hierarchical orders, ranking hierarchical orders and consistency checking. The last three steps are carried out layer by layer. Analytic Hierarchy Process has the following four characteristics:
1) The principle is simple. Analytic Hierarchy Process (AHP) is based on experimental psychology and matrix theory, which is easy to accept. Its principle is clear, concise and easy to use; There are few requirements for quantitative information.
2) Clear structure. The idea of solving problems is to divide complex problems into several simple problems with structure and hierarchy, and then solve the simple problems.
3) Combination of qualitative and quantitative. AHP method is to determine the judgment matrix through expert investigation (such as Delphi method), determine the weight of each index element through rigorous quantitative method, and finally make comprehensive evaluation and optimization.
4) Analytic Hierarchy Process (AHP) is suitable for determining the weight of each factor in the multi-factor and multi-level dynamic system of investment environment. It can effectively objectify the subjective judgment of experts in system engineering, and can solve problems that can not be dealt with only by quantitative methods. It is widely used in decision analysis of social and economic systems.
(4) Principal component analysis
Principal component analysis (PCA) is "transforming a given set of related variables into another set of irrelevant variables through linear transformation in mathematics". It arranges the new variables in the order of decreasing variance, reduces the interference among the factors in the index system, facilitates the search for the leading factors and simplifies the evaluation process.
Principal component analysis is to reduce the dimension of high-dimensional variable space under the principle of ensuring the most complete data, that is, to optimize the comprehensive simplification of multivariate data. Its advantage is that it can obtain objective weight and avoid the deviation caused by human factors. The disadvantage is that economic factors are not considered in pure mathematical calculation, and it is difficult to make a reasonable explanation of the economic significance of principal components according to objective reality; Ignoring the importance of the index itself, the weight of the index is too different from the expectation, and the original index information extracted by it is the difference information of the data, not the more important information such as the meaning and importance of the index.
The optimization method of principal component comprehensive evaluation is a kind of relative evaluation optimization, and its evaluation optimization standard (evaluation function) is related to the selection of samples. The number, addition and deletion of evaluation optimization units will affect the evaluation optimization conclusion. Moreover, the optimization method of principal component comprehensive evaluation can not eliminate the overlapping information of indicators. It is easy to be influenced by the overlapping of indicators, which leads to the close relationship between the optimal result of comprehensive evaluation and the correlation structure of indicators.
(5) Optimization method of fuzzy comprehensive evaluation
The fuzzy comprehensive evaluation and optimization method founded by American scientist L.A.Zadch in 1960s is a comprehensive evaluation and optimization method based on fuzzy mathematics and its corresponding fuzzy statistics, which comprehensively considers many factors that affect something and quantitatively describes some factors that are unclear and difficult to quantify.
The optimization method of fuzzy comprehensive evaluation includes: ① defining the comprehensive evaluation optimization system (with P indexes), that is, the factor universe U; (2) define a comment level domain v (with m comment levels); ③ Determine the index weight w; (4) establishing a fuzzy relation matrix r; ⑤ Calculate the fuzzy comprehensive value b; ⑥ Fuzzy comprehensive evaluation and optimization, that is, the evaluation and optimization results are transformed into a sortable form, and the comprehensive evaluation and optimization are sorted or classified, that is, fuzzy category recognition.
Relevant experts constantly develop fuzzy comprehensive evaluation optimization methods in their application practice, and their application fields are also expanding, and various methods combined with fuzzy comprehensive evaluation optimization methods are also expanding in many fields. Because the domain is very complex, the models will vary widely, leading to more and more complex models.
The advantage of fuzzy comprehensive evaluation and optimization method is that when dealing with the evaluation and optimization of multi-level complex problems and multi-factor comprehensive judgment, the influence of various factors on the whole is comprehensively considered, and people's subjective experience can be reflected by objective figures.
However, when quantifying some ambiguous and difficult-to-quantify factors, the information carried by the factors may be lost, and the quantized factors are very different from those before quantification.
(6) Cluster analysis combined with principal component analysis is used to evaluate the optimization method.
Firstly, cluster analysis is carried out to distinguish the influence of the evaluated preference object, and then the preference value of the evaluated person is calculated by principal component analysis, and the intra-class and inter-class ranking is carried out according to the preference value. This method takes into account the specific objectives of evaluation and optimization, so that the correlation between evaluation and optimization indicators can be eliminated to the greatest extent, and at the same time, important significant indicators are considered, so as to minimize the subjective error in determining the weight of indicators, thus improving the effectiveness of evaluation and optimization results.
However, this method must set the decisive evaluation optimization factors in advance and cluster according to these factors, which requires high quantification of indicators.
(7) Evaluation and optimization method combining analytic hierarchy process with fuzzy comprehensive evaluation and optimization.
Using analytic hierarchy process to determine the weight of each factor. After getting the weight, the fuzzy mathematics method is used to calculate the evaluation optimization value of each factor, and then the comprehensive score is obtained. The combination of analytic hierarchy process and fuzzy comprehensive evaluation optimization method is very suitable for comprehensive evaluation of venture capital environment in the region. This method considers the multi-level and complexity of venture capital environmental factors, and is not suitable for the evaluation and optimization of venture capital environment in the whole region.
(8) Entropy weight method
Entropy (meaning changing capacity) was put forward by the German physicist R.J.E Clausius when he studied the thermal cycle in 1864. 1948, N. Wiener and C. E. Shannon founded information theory. Shennong called the uncertainty in the process of information source transmitting signals information entropy, which showed the relationship between selection and uncertainty and random practice, and solved the problem of quantitative description of information. The increase of entropy means the loss of information; The more orderly the system, the smaller the entropy and the greater the information. The greater the entropy, the smaller the information. 1967, Thiel put forward theil index for the first time when he studied the income gap between countries. It is obtained by subtracting the information entropy value h from the constant LgN. This method takes the difference of information reflected by entropy in different probability events in information theory and system theory as the theoretical basis of weight setting, that is, for a system composed of multiple index values, the greater the dispersion of index values, the smaller the information entropy value of indicators, and the greater the importance of reflection.
Entropy weight method is a mathematical method to calculate the comprehensive index according to the information provided by various factors. It is an objective and comprehensive weight determination method, based on the amount of information transmitted by each index to decision makers.