The paper analyzes the factors that affect the amount of effort required to create a mobile application and its cost. It is established that the main factors of influence are the design of the application, its functionality, the type of mobile platform, the availability and level of testing and support, as well as the individual characteristics of the developer. Based on the analysis of information sources, the main methods and approaches to forecasting the cost of software products are identified, which include the COCOMO model, Price-to-win method, expert evaluation, algorithmic methods and the method of analogies. It is proposed to consider the method of analogies as a tool that allows you to make predictions about the cost of resources required for the successful implementation of IT projects based on the experience of similar projects. It is proved that the advantages of this method are the simplicity of its implementation and the clarity of the results obtained, which follows from the practical orientation of this tool. Among the limitations of the method of analogy is the mandatory need for reliable data relating to similar projects, as well as the difficulty of taking into account unspecified indicators. Taking into account the mentioned limitations of the method of analogies and on the basis of the analysis of scientific sources the possible directions of its optimization are determined. Thus, among the ways to improve the effectiveness of this method are those aimed at optimizing the project selection process, the data for which are used as a basis for forecasting. Attempts to improve the method of analogies by including parameters that were previously ignored by this technique seem promising. This in turn can lead to an expansion of the scope of the method of analogies and increase the accuracy of forecasts. As prospects for further research, the need to continue research in the field of optimization of the method of analogies with the subsequent practical verification of theoretical positions on the data of real projects.
Tuning of COCOMO Model Parameters by using Bee Colony Optimization
