Abstract
Venture capital plays a significant role in economy. Different types of companies involved in venture capital activities may face difficulty in financing. This study aimed to investigate the determinants of venture capital in the companies listed on the Tehran Stock Exchange.
 The present study was applied in terms of purpose and In terms of the descriptive research method, it is a survey with a comparative causal approach. A number of 100 companies listed on the Tehran Stock Exchange in 2014 to 2019 were selected as a sample. Data were collected from the databases of the Central Bank of Iran, Statistics Center of Iran, Codal website, and WDI website and Tehran Stock Exchange website, and then analyzed using Excel, 9Eviews, and Matlab software. In this study, the determinants of venture capital on the Tehran Stock Exchange (capital volume, export volume, company tax, disclosure index, rule of law index, industry diversity, diversity of company life cycle stages, type of property, number of divisions and subsidiaries, company capital, company, size and company age) By studying the research literature and using neural network and fuzzy logic, a predictive model was created to determine the investment risk based on the determined indicators.. The classification model was presented using neural network without SOM supervision and fuzzy logic flexibility. The evaluation parameters of design, accuracy of the proposed ensemble classifier, mean error of the squares, and coefficient of explanation. The simulation results indicated that the performance of the proposed design compared to the classification by MLP neural network method was improved and the presented model based on the training provided predicted the status of capital with good accuracy.
Introduction
Venture capital and its role in economy has been raised since the 1990s and includes not only financial capital but also non-financial capital. (Gompers and Lerner, 1998). Venture capital is described as a means of providing capital to the companies which may not use independent financial instruments and thus require external financing. Venture capital is a kind of active capital in the market in which some activities like monitoring and influencing the company's strategic decisions are performed using salary control and board seats. The abilities of an individual to participate in venture capital to add more value is reflected in the duration of the capital (Hain et al., 2018).
Materials and Method
The data set used in this simulation included capitals made by the companies listed on the stock exchange in 2014 to 2019, with a record of 500 capitals selected as a sample. Such data included two parts: training data and testing data to evaluate the accuracy of the model in forecasting. Of all the devices, 350 capitol records were used as training data and 150 capital data were as test datasets to evaluate project performance. Each capital record was assigned a numerical value as the amount of venture capital. The present study was applied in terms of purpose. The purpose of applied research is to develop applied knowledge in a specific field. In addition, the present study was descriptive-correlational in terms of method and nature. This study aimed to determine the relationship between variables. For this purpose, appropriate indicators were obtained based on the scale of measurement of variables. The study was conducted in form of inductive deduction and its information was expost facto. The present study was a thematic area in the field of finance with a focus on venture capital issues. Tehran Stock Exchange was selected as the spatial domain of the present study. Data were collected from the databases of the Central Bank of Iran, Statistics Center of Iran, Codal website, Tehran Stock Exchange website, and then analyzed using Excel, 9Eviews, and Matlab software. The research process was as follows: in the framework of developing economies, focusing on Iran, the determinants of venture capital on the Tehran Stock Exchange were identified. Then, a predictive model was created using neural network and fuzzy logic to determine venture capital based on extractive indicators. capital volume, export volume, company tax, disclosure index, rule of law index, industry diversity, diversity of life cycle stages of the company, type of property, number of divisions and subsidiaries, company capital, company size and age of the company were twelve factors determining venture capital on the Tehran Stock Exchange.
Hypothesis
1:  The investment volume of mergers and acquisitions has an effect on venture investment in the Tehran Stock Exchange.
2:  Exports have an effect on venture investment in Tehran Stock Exchange.
3:  Company tax has an effect on venture investment in Tehran Stock Exchange.
4: Disclosure index has an effect on venture investment in Tehran Stock Exchange.
5: The rule of law index has an effect on venture investment in the Tehran Stock Exchange.
6: Diversification of industry has an effect on venture investment in Tehran Stock Exchange.
7:  Diversification of different stages of the company's life cycle has an effect on venture investment in the Tehran Stock Exchange.
8: The type of ownership has an effect on venture investment in Tehran Stock Exchange.
9: The number of sections and subsidiaries has an effect on venture investment in Tehran Stock Exchange.
10: The company investment has an effect on venture investment in Tehran Stock Exchange.
11: Company size has an effect on venture investment in Tehran Stock Exchange.
12: The age of company has an effect on venture investment in the Tehran Stock Exchange.
Discussion and Results
 The design evaluation parameter was the mean square error (MSE) and the explanation index (R2) of the proposed ensemble classifier
In order to compare the proposed design with existing tools, it was necessary to determine the accuracy of these tools on their data set, so in this section, the MLP neural network was used first to predict venture capital. At this step, 12 features were proposed for corporate capital records as input neurons to the neural network. The output of the network was the amount of venture capital. Here, the neural network had the task to provide conditions in the model training phase by determining the weights, so that the predicted amount of venture capital had the least difference with their actual amount.
After training, the neural network was evaluated to test performance with testing data.
Conclusion
In the proposed design, to predict the initial clustering of data by SOM method, to calculate the centers of clusters, the innovative formula was used and combined with fuzzy logic of the second type was used.
In the current design, the researchers used the SOM neural network, which is an unsupervised learning method, along with logic flexibility, to decide on the amount of venture capital. The SOM network clusters the observed data and determines the centers of the clusters, then assigns a probabilistic value to each cluster in the range of zero and one, depending on the sample density in each cluster. After that, a model for labeling new samples based on the distance factor from the centers of the clusters was presented, which based on the second type fuzzy logic, the probability of each sample belonging to each class is determined. The simulation results indicate that the performance of the proposed design has improved compared to the classification by MLP neural network method and the proposed model based on the training provided to the model, has predicted the status of capitals with good accuracy.