Текст книги "Effective Methods and Transportation Processes Management Models at the Railway Transport. Textbook"

2.6 Data processing statistical models

Statistical methods are a class of methods of formalized representation of systems that are used in cases where a preliminary analysis of a problem situation shows that it cannot be represented in the form of a well-organized system, then it is recommended to present the situation in the form of a poorly organized, or diffuse, system and turn primarily to statistical methods.

The basis of these representations is the display of phenomena and processes with the help of random (stochastic) events and their behavior, which are described by the corresponding probabilistic (statistical) characteristics and statistical patterns.

The term stochastic clarifies the concept of random, which in the ordinary sense is usually associated with the absence of causes for the occurrence of events, with the appearance of not only repetitive and obeying some laws, but also single events; The processes displayed by statistical regularities must be rigidly connected with predetermined, certain causes, and randomness means that they may or may not appear in the presence of a given set of causes.

Statistical mappings of the system in the general case (by analogy with analytical) can be represented as a «blurred» point (blurred area) in n-dimensional space, into which the system (its properties taken into account in the model) is translated by the operator F [SX]. A «blurred» point should be understood as an area that characterizes the movement of the system (its behavior); in this case, the boundaries of the region are given with a certain probability p (the probability of the event is understood p (A) = t / n, where t is the number of occurrences of the event A, n is the total number of experiments; if at n -> ∞ (t / n) -> const.), i.e. «blurred», and the movement of the point is described by some random function.

Statistical regularities can be represented in the form of discrete random variables and their probabilities or in the form of continuous dependencies of the distribution of events and processes.

Practical application is mainly one-dimensional distributions, which is associated with the complexity of obtaining statistical patterns and proving the adequacy of their application for specific applications, which is based on the concept of sampling.

A sample is understood as a part of the studied set of phenomena, on the basis of the study of which statistical patterns are obtained that are inherent in the entire population and extended to it with some probability.

In order for the patterns obtained in the study of the sample to be extended to the entire population, the sample must be representative (representative), i.e. have certain qualitative and quantitative characteristics. Qualitative characteristics are related to the content aspect of the sample, i.e. with determining whether the elements included in the sample are elements of the population under study, whether these elements are correctly selected from the standpoint of the purpose of the study (from this point of view, the sample can be random, directed or mixed). The quantitative characteristics of the representativeness of the sample are related to the determination of the sample size sufficient to draw conclusions about the population as a whole on the basis of its study; A decrease in sample size can be obtained on the basis of ergodic properties, i.e. by increasing the duration of statistical tests (in most practical cases, the question of the quantitative characteristics of the sample is the subject of a special study).

On the basis of statistical representations, a number of mathematical theories are developed:

• probability theory and mathematical statistics, combining various methods of statistical analysis (regression, variance, correlation, factorial, etc.);

• the theory of statistical tests, which is based on the Monte Carlo method, and the development is the theory of statistical simulation modeling;

• the theory of putting forward and testing statistical hypotheses, which arose to evaluate the processes of signal transmission at a distance and is based on the general theory of statistical decision functions of A. Wald. A special case of hypothesis theory, important for systems theory, is the Bayesian approach to the study of information transfer processes in the processes of communication, learning and other situations in organizational systems;

• the theory of potential noise immunity, the beginnings of which were laid by the works of V.A. Kotelnikov, carried out independently of the theory of decision functions;

• generalizing the last two directions of the theory of statistical decisions, within the framework of which, in turn, a number of interesting and useful directions for practice have arisen.

These areas are mostly theoretical and applied in nature and arose from the needs of practice. However, there are a number of disciplines that are more pronounced applied in nature. Among them are statistical radio engineering, statistical theory of pattern recognition, economic statistics, queuing theory, as well as stochastic programming, new sections of game theory, etc.

The expansion of the possibilities of displaying complex systems and processes in comparison with analytical methods can be explained by the fact that in the case of using statistical representations, the process of setting the problem, as it were, is partially replaced by statistical studies that allow, without identifying all deterministic relationships between the objects (events) under study or the components of a complex system taken into account, on the basis of a sample study (study of a representative sample) to obtain statistical patterns and extend them to the behavior of the system as a whole.

However, it is not always possible to obtain statistical patterns, it is not always possible to determine a representative sample, and the legitimacy of applying statistical patterns is proved. If it is not possible to prove the representativeness of the sample or it takes an unacceptably long time, then the use of statistical methods can lead to incorrect results.

In such cases, it is advisable to turn to the methods united under the general name – the methods of discrete mathematics, which help to develop modeling languages, models and methods for the gradual formalization of the decision-making process.

Statistical and set-theoretic methods initiated the emergence of the theory of «fuzzy» sets J1. Zadeh, which, in turn, was the beginning of the development of a new direction – the theory of fuzzy formalizations.

Stochastic programming (SP) is a branch of optimization theory devoted to the study and solution of extremal problems in which there is no accurate information about the values of the objective function and/or constraints. In such problems, the objective function and/or constraints usually depend on one or more random parameters.

The term stochastic programming appeared in the early 50s of the XX century, when Danzig, Charns and Cooper began to analyze linear programming problems with random coefficients that arise when planning in situations with uncertainty and risk.

When solving SP problems, it is impossible to do with deterministic methods and special stochastic procedures have to be used. In SP problems, some characteristic of a random function, for example, its mathematical expectation, is usually subject to maximization or minimization. At the same time, in some formulations of the tasks of the joint venture, it is allowed to fulfill the restriction in the form of equality (or inequality) with some positive probability.

Strategic management (strategic management) is a term that has become widely used in economics and management of enterprises and organizations in the 60—70s of the XX century.

The first planning systems in world practice (1900—1950) were based on the preparation of annual financial estimates for items of expenditure for various purposes. Their main task was to manage costs. During this period of development of economic relations, the implementation of financial planning was considered a sufficient condition for the effective functioning of economic entities. In the development of the methodology of strategic management (SM), this stage is called «management based on control over execution», while the possible reaction of organizations to changes was determined after the events.

In the 1950s and 1970s, with the acceleration of economic development and competition, companies can no longer rely on budgeting as a system of preparation for future competitive problems. In order to increase competitiveness in the new environment, they switched to long-term planning, which quickly proved its usefulness and was adopted by most large and significant numbers of medium-sized firms. At this stage, the methodology of the management process was based on the basis of «extrapolation of past trends».

The founder of SM is considered to be I. Ansoff, who proposed a new concept of strategic planning, based not on the forecast of probabilistic trends in the development of the organization, but on an entrepreneurial approach.

With the growth of crisis phenomena (the energy crisis of 1970), the tightening of competition, forecasts based on extrapolation ceased to meet the requirements of a dynamically changing external environment. Financial and long-term planning was integrated into strategic planning, the purpose of which was to determine the future market position so that the company could adequately respond to its changes. The new method is called «management based on anticipation of change.»

In the 1970s and 1990s, Western firms moved from strategic planning to the SM of their activities, which is also called market, while emphasizing the external orientation of the organization’s management. This approach to management allows business entities to move from a reactive form of management (making management decisions as a reaction to current problems) to management based on analysis and forecasts.

This allows not only to respond to changes in the external environment, but also to create them, to influence them. The use of the term «strategic management» instead of the term «strategic planning» implies an emphasis on an entrepreneurial approach and taking into account the interests of the organization’s employees.

The methodological principle of modern strategic management is to build a strategy not from the past to the present, but from the future through the past to the present.

Thus, the emergence of the methodology of strategic management, as well as innovation management, is usually considered from the standpoint of the evolution of planning systems as a reaction of economic entities to the complication of external business conditions.

The term «strategy» is borrowed from the military lexicon and is interpreted ambiguously in various definitions: in some – as a definition of goals, i.e. promising (strategic) areas of the enterprise, taking into account its purpose (mission); in others – as a display of goals, in the form of a plan (i.e., indicating the deadlines, performers and other conditions for the implementation of goals); thirdly, as finding ways to achieve goals; Fourthly, as a choice of methods, a set of rules for decision-making, or even a choice of means to achieve goals.

In other words, the term «strategy» is used at all stages of decision-making – from the formulation of goals to the choice of methods and means of their implementation.

There are recommendations for the definition of a list of stages based on an analysis of the basic principles and conditions of the CM.

So, B. Karloff considers five conditions necessary for the implementation of strategic management:

1. Ability to simulate a situation based on a holistic view.

2. The ability to identify the need for change (taking into account the variety of variables – from the efficiency of production costs to the differentiation of the product range, including the assessment of product quality, risk, etc.).

3. Ability to develop a change strategy.

4. Ability to use reliable methods in the course of change.

5. Ability to put strategy into practice.

These conditions can also be considered as stages of strategic management.

At all stages of SM, approaches, methods and techniques of system analysis can be used. And at the same time, the achievements, approaches developed in the theory of SM independently of the theory of systems, practical experience, reflected in the specific recommendations contained in the works on SM, are useful for the development of system analysis.

The main ones are as follows.

Mission and strategic goals. The primary task of the SM is a stable presence in promising and stable markets with a diverse and continuously improving (taking into account the needs of the market) assortment that is competitive in price, quality and methods of promotion.

Mission is a concept that is sometimes used as a definition of a business goal or concept. Models are also considered, according to which the sequence «mission – concept – goal» takes place. At the same time, researchers note the complexity of working with the concept of purpose, and therefore propose concepts that replace this term with others, in relation to the choice of the main activities of the enterprise.

At present, the use of the concept of key competence is becoming more widespread, which is more convenient to interpret in practice than the rather complex concept of goal and makes an important contribution to the approximation of the theory of goal formation to the practice of managing specific enterprises and organizations.

Key competency. D. Campbell defines competence as a property or series of properties inherent in all or most companies in the industry. Only by possessing them, the organization can participate in entrepreneurial activities. A key (core) competence is a distinctive feature, property or a number of properties specific to a particular organization, which allows you to produce goods of above-average quality and use your resources and competencies more efficiently.

The main properties of the core competence are: potential access to different markets; adding significant use value to the final product; the ability to use it only within a certain business system; indispensability – cannot be replaced by another competence, etc.

The identification of key competencies is considered as one of the main components of the organization’s success, an integral part of the SM, since it is thanks to it that the company is able to maintain its position in the market and defeat competitors. The definition of competitive advantage acts as the main goal of the business strategy. The core competence determines the company’s leadership in the market, and in order not to lose this leadership, it is necessary to protect and improve its advantages all the time. In order to correctly assess the situation and effectively adjust your actions, it is necessary to understand that the key competence is formed from a number of competencies that form the basis of the enterprise’s activities. To analyze key competencies, the direction of their management (Competence Management) is being developed.

Basic principles of the formation of key competence. When determining the key competence, two main categories are taken into account: resources and competencies that form the internal conditions of the organization, which are a set of production and technological, financial and economic, socio-cultural, organizational, technical and administrative conditions. Simulation of internal conditions lays the fundamental basis for further analysis. Together with external conditions (economic, political, legal, socio-cultural, technological), they determine the set of resources available to the organization, as well as the form and content of its business processes, as a result of which a product that satisfies social needs appears.

Any strategy should be based on competitive advantages. They allow the company to have a profitability above the average for firms in a given industry or a given market segment (which is ensured by higher efficiency in the use of resources) and gain a strong position in the market. When developing a competition strategy, it is necessary, on the one hand, to have a clear idea of the strengths and weaknesses of the enterprise, its position in the market, and on the other hand, to understand the structure of the national economy as a whole and the industry in which the enterprise operates. In the historical aspect, the theory of competitive advantage has replaced the theory of comparative advantage. The comparative advantages underlying the competitiveness of a country or firm are determined by the availability and use of abundant factors of production, such as labor and raw materials, capital, infrastructure, etc. But with the development of technological innovation and the globalization of business, the structure of international competition is changing, and comparative advantage is being replaced by a new paradigm – competitive advantage. This means the following:

Advantages change under the influence of the innovation process (production technologies, management methods, methods of delivery and marketing of products, etc.). Therefore, in order to maintain competitive advantages, constant innovation is required.

The globalization of business forces companies to take into account national and international interests. The state, the territory are considered as the basis of the company’s strategy, and not only as the place where the company operates.

The choice of a core competence is a complex process that includes consideration of many aspects in a complex. First of all, the firm must analyze five competitive forces: potential market participants, buyers, suppliers, substitute goods, competitors. Key competencies can take a variety of forms, depending on the specifics of the industry, product and market. When defining key competencies, it is important to focus on the needs of consumers and make sure that these benefits are perceived by them as such.

The main requirement is that the difference from competitors must be real, expressive, significant.

Competitive advantages are not eternal, they are won and maintained only with the continuous improvement of all areas of activity, which is a laborious and, as a rule, expensive process. The possibility of maintaining competitive advantages depends on a number of factors: 1. Sources of competitive advantages. 2. Evidence of sources of competitive advantage. 3. Innovation. 4. Giving up an existing competitive advantage to acquire a new one.

STEP and SWOT analysis are models for analyzing social (Social), technological (Technological), economic (Economical), political (Political) in relation to strengths (Strength) and weaknesses (Weakness) sides, opportunities (Opportunities) and threats (Threats).

SWOT analysis is a matrix analysis of the organization’s activities, integrating the study of the company’s capabilities in the context of environmental challenges and business responses. SWOT is an abbreviation: strong – strong; weak – weak; opportunity– opportunity; threat – threat.

In the theory of system analysis, STEP and SWOT analysis correspond to two stages of the system analysis methodology – the stage of forming the structure of goals and functions and the stage of evaluating the components of this structure.

STEP analysis can be considered as one of the structuring techniques, recommending the definition of top-level sub-goals based on the analysis of social, technological, economic and political factors. At the same time, in order to ensure the completeness of the identification of factors, it is advisable to take into account one of the important laws of systems theory – the pattern of communicativeness, i.e. an analysis of the factors of the supersystem, subordinate systems, the current environment and the system itself is carried out.

A SWOT analysis defines criteria for a qualitative assessment of factors in terms of strengths, weaknesses, opportunities, and threats. The application of the assessments recommended in the PATTERN methodology expands the composition of the evaluation criteria. At the same time, the estimates recommended in the SWOT analysis are a refinement of the estimates recommended in the PATTERN, and they are useful for use in other methods of system analysis.

Estimates in SWOT analysis are sometimes presented in the form of a two-dimensional matrix with SW and OT axes, which allows you to get more generalized estimates: SO («strong», i.e. great opportunities), ST (strong threats), WO (weak opportunities), WT (weak threats).

Models for generating strategies based on BCG, Ansoff, Porter matrices, or portfolio analysis methods. The main method of portfolio analysis is the construction of two-dimensional matrices, on one axis of which the values of internal factors are recorded (assessment of the competitiveness of the organization’s divisions), on the other – external (assessment of market development prospects). With the help of these matrices, products (or other business units) can be compared with one another according to a number of criteria (sales rates, competitive position, life cycle stage, market share, attractiveness, etc.).

The main advantages of portfolio analysis are the possibility of some logical structuring and visibility of the display of strategic problems, the relative simplicity of presenting the results using qualitative analysis criteria.

A further stage in the development of portfolio analysis was the work of Bruce Henderson, founder of the Boston Consulting Group (BCG). The axes of the first BCG matrix were market growth/market share. For greater clarity, the elements of the matrix are given specific names

«Stars» are enterprises that have won a large market share in growing sectors of the economy, «Cash Cows» – in mature industries, «Dogs» – enterprises that have won a small market share in industries experiencing stagnation. A «question mark» (or sometimes referred to as «Calves») describes businesses that have gained a small market share in fast-growing industries.

Subsequently, based on the ideas of the BCG matrix, three-dimensional matrices were proposed, the axes of which form complex indicators: the attractiveness of the market, the competitive position of the enterprise, the competitiveness of the goods.

Formalized models and automated procedures for their implementation are developed based on the ideas of matrices.