Cognitive Map Reflecting the Results of Simulation Modelling and Neural Network Analysis

Selection of appropriate subjects of study, which most accurately illustrate the characteristics and regularities under study, is among the crucial objectives of scientific research. Wrong selection of subjects leads to the negative research results after a long collection of statistical data and its complex and time-consuming analysis, rendering all efforts vain and preventing from solving the problem posed. In this case, the scholar is either forced to repeat research with other subjects of study, which implies an excessive consumption of research resources, or abandons the idea at all, which hinders the advancement of science.

In order to simplify and speed up the process of selecting the subjects of study, as well as to increase the probability of the right selection of subjects to be studied in each specific case, we have provided the possibility for the express data analysis in our dataset. Depending on the problem posed, the logic of the research, and the original author's idea, different combinations of countries and indicators can be selected, and the common links between them can be observed. All calculations are done automatically in real time, making it possible to conduct the necessary number of tests and select those particular countries and indicators that are best suited for each particular research.

The dataset allows for the automatic simulation modelling and neural network analysis of data using the correlation analysis method. For clarity, the results are presented as a cognitive map, a mathematical graph which represents the set of elements being studied and the relationships between them. The advantage of the cognitive map is that it shows a multiple rather than a pairwise correlation, thereby exposing the interconnections between all indicators. This allows, inter alia, excluding duplicate variables from the research.

When the cognitive map was constructed, indicators from the category of state regulation of corporate social and environmental responsibility, as well as indicators from the category of market-based management of corporate social and environmental responsibility were visually separated from each other for the sake of convenience: they were isolated and indicated by a different color on the cognitive map in the selection of indicators. This further simplifies the selection of indicators according to the authors' classification.

The algorithm of simulation modelling and neural network analysis is as follows:

First, countries are selected from both the general list of countries of the world and the ready-made templates;
Indicators are then selected according to the classification. One process of simulation modelling and neural network analysis can link only one indicator of sustainable development and combating climate change with any multitude of state and/or market-based management indicators;
Cross-correlation is automatically determined for selected countries and indicators;
A cognitive map is ultimately constructed based on the results of simulation modelling and neural network analysis. The values of their correlation coefficients are shown on the map at the intersection of every two indicators.
In case of need, the algorithm described above can be repeated as many times as necessary until those particular countries and indicators for which there is a correlation required for this research are identified.