At present, no sphere of society can be managed without forecast as a method of foresight. Forecasting is used in various fields: in economics, management, sports, industry, etc. It is possible to draw preliminary conclusions about various processes, phenomena, reactions and operations using extrapolation and trend.
The essence of forecasting
Socio-economic forecasting is an important scientific factor in the strategies and tactics of social development. Therefore, research questions and forecasting methods are quite relevant. The problem of forecast relevance is also determined by the degree of risk (for example, financial risks) in decision-making in areas such as regional management, inventory control, production planning, financial planning, etc.
Forecast results are used to support decision making. Therefore, the nature of the decisions is determined by the majority of the desired characteristics of the systemforecasting. Studying this problem should help answer questions about what to predict, what form the forecast should take, what time elements should be included, what is the required accuracy of the forecast.
The uncertainty of the external environment in the future and the lack of information about the state of the object under the influence of various external and internal conditions make the task of forecasting quite difficult, and the process itself may not always fit into a certain algorithm. This leads researchers to look for new ways to solve problems using probability theory and mathematical statistics, combinatorial theory and nonlinear dynamics, etc.
The development of work on issues related to forecasts is carried out in such main directions as:
- intensification of theoretical and applied research of several groups of methods that meet the requirements of various objects and types of forecasting;
- development and implementation in practice of special methods and procedures for the use of various methodological techniques during a particular study;
- pathfinding and algorithmic presentation of forecasting methods, as well as their implementation using computers.
Classification problem
The issue of studying and categorizing forecasting methods is very relevant, due to the possibilities of its application in accordance with the required type of forecasting object and forecast form. It is necessary to study the theoretical andmethodological aspects of forecasting, determine the role of forecasting in the object management system. This is important for clarifying the tasks, functions and principles of forecasting, for organizing the classification functions of forecasting, and finding out its essence. Another task is to characterize and analyze current forecasting methods, to analyze the possibilities of using various forecasting methods in solving various types of practical problems.
Definition
Forecasting is defined as a method that uses theoretical and practical steps to develop forecasts. This definition is general and allows us to understand this term quite broadly: from simple extrapolation calculations to complex multi-stage expert research procedures.
Basic concepts
There are some basic concepts within the research subject.
The forecasting stage is a part of the forecast development process, which has certain tasks, methods and results. The division into stages is connected with the features of the construction of the process, which includes:
- systematic description of the forecasting object;
- data collection;
- simulation;
- forecast.
The predictor model is a model of the predictor object that provides information about the possible future states of the predictor object and/or how and when they might be realized.
Forecasting methods representa set of special rules and methods (one or more) that ensure the development of the forecast.
Forecasting system is a system of methods that operate in accordance with the basic principles of forecasting. Implementation methods are a group of experts, a set of programs, etc. Prediction systems can be automated and non-automated.
The object of forecasting is a process, system or phenomenon, the state of which is determined by the forecast. The forecast variable object is a quantitative characteristic of the forecast object, which is taken as a variable related to the time range of the forecast.
Forecasting technique is a set of special rules and methods used to develop specific forecasts.
Forecast can be simple or complex. A simple forecast is a method that cannot be divided into simpler forecasting methods. Complex forecasting is a method consisting of a coherent combination of several simple methods.
Consistency of methods
Currently, the problem of choosing a forecast method has several criteria, this process is poorly designed and not fully structured. The fundamental principle for solving such a problem is the principle of consistency.
The system approach allows you to discover and implement the principle of consistency. It is universal and corresponds to the method of analysis and study of any complex systems.
BWithin the framework of this approach, the properties, structure and functions of objects, phenomena and processes as a whole are studied by representing them as systems with all complex interelement relationships, the mutual influence of elements on the system and the environment, as well as the influence of the system on structural elements.
The consistency of forecasting methods and models is understood as the possibility of their joint use, which allows making a consistent and consistent forecast of the development of an object. This method is based on the study of current and future trends in regularity, according to the specified parameters, available resources, identified needs and their dynamics.
Methodology
The forecasting system includes a certain order of using the model for the formation of a comprehensive forecast of the object or phenomenon under study. This method helps to define the forecasting methodology. It includes a set of forecasting models, methods and methods of calculation.
The systematic method of research is especially important for solving complex problems. The need for a systematic approach to forecasting follows from the peculiarities of the development of science and technology. A large number of elements, objects of different types, complex relationships between them and the behavior of an object in the external environment led to the creation of large technical and industrial (organizational-economic) systems.
Classification basics
Currently, along with a significant number of published forecasting methods, there aremany ways to classify them. The main objectives of the classification of forecasting methods:
- supporting the research and analysis process;
- Supporting the selection process for developing object forecasts.
Today it is difficult to offer a general classification that is equally relevant to these two purposes.
Prediction methods can be classified according to several attributes. One of the most important classification criteria is the degree of formalization, which quite fully covers forecasting methods.
In general, the classification is open as it provides the ability to increase the number of elements in the levels and increase the number of levels through further fragmentation and specification of the final level elements.
Another approach to definition
According to a more precise definition of the concept of forecasting, types of forecasts, it is a set of methods and ways of thinking that make it possible to judge its (object) future development. It is based on the analysis of historical data, exogenous (external) and endogenous (internal) relations of the object of forecasting, as well as their measurement within the framework of this phenomenon or process.
Classification criteria are also the unity of the classification attribute at each level; disjunctive classification of one section; and openness of the classification scheme.
In turn, each level in the scheme is determined by its own classification criterion: the degree of formalization, the general principle of action; way to get a forecast.
Classification of methods
From the point of view of the general approach, many forecasting methods aimed at solving applied problems of analyzing the state of an object and forecasting its current development can be represented within the following classification.
The main types of forecasting, in accordance with the degree of formalization, can be intuitive and formalized.
Intuitives can be individual and collective.
Individual, in turn, are divided into interviews, questionnaires and processing of analytical hierarchies. Collective methods include the Delphi method, brainstorming, expert commission, script building.
Formalized methods can be mathematical, system-structural, associative. Also in this category are methods of promoting information.
Mathematical methods are divided into two categories: statistical and extrapolar.
The first category is represented by correlation analysis, regression analysis, time series models, adaptive models.
The second category is represented by moving average and exponential smoothing.
Mathematical methods also include combinational methods.
System-structural methods are represented by morphological analysis, functional-hierarchical modeling, network modeling and matrix modeling.
Associative methods include simulation, historical analogy, data mining.
The types of forecasting includeSee also methods for promoting information presented by the analysis of the flow of publications, the significance of the invention and the analysis of patents.
Characterization of intuitive methods
Expert (intuitive, heuristic) types of forecasting are based on information received from professional experts as a result of systematic processes of identification and synthesis. These methods require experts to have deep theoretical knowledge and practical skills in collecting and synthesizing all available information about the forecasting object.
Intuition (unstructured knowledge) helps specialists to identify trends in the development of the forecasting object without any basic information about it. For example, the forecast of demand for new goods and services, the effectiveness of innovation, the end of economic reform, world prices for energy products, metals (non-ferrous and precious) and even currencies.
Such types and methods of forecasting as expert ones are usually used in the following cases:
- when it is impossible to consider the influence of many factors due to the significant complexity of the forecasting object;
- when there is a high degree of uncertainty in the available information in the forecast base.
Thus, intuitive methods are used when the predicted object is either too simple, or too complex and unpredictable, so that it is almost impossible to analyze the influence of many factors analytically.
Collective methods of expert judgments are based on the fact that the collectiveconsciousness provides a higher accuracy of results. In addition, when processing the results obtained, unproductive (extraordinary, abstract) ideas may arise.
Characteristics of formalized methods
Formalized (factual) types of forecasting are based on the actual and available information of the forecasting object and its past development. They are used in cases where information about the forecasting object is mainly quantitative, and the influence of various factors can be explained by mathematical formulas.
The advantage of this group of methods is the objectivity of the forecast, expanding the possibility of considering various options. However, in the methodology of formalization, many aspects remain outside the analysis. Thus, the greater the degree of formalization, the poorer the model.
Until recently, the statistical method was the main method in the practice of forecasting. This is mainly due to the fact that statistical methods rely on technique analysis, development and application practices that have a fairly long history.
The process based on statistical types of planning and forecasting is divided into two stages. First, a generalization of the data collected for a certain period of time, and the creation of a process model based on this generalization. The model is described as analytical expressions of a development trend (extrapolation trend) or as a functional dependence on one or more argument factors (regression equations). Any kind of predictive model shouldinclude the choice of the form of the equation that describes the dynamics of the phenomenon, the relationship and the assessment of its parameters using a specific method.
The second stage is the forecast itself. At this stage, based on various patterns, the expected value of the projected pattern, size or characteristic is determined.
Of course, the results obtained cannot be considered as a final conclusion. During their evaluation and use of factors, conditions and constraints, all factors that were not involved in the specification and construction of the model should be taken into account. Their adjustment should be carried out in accordance with the expected change in the circumstances of their formation.
Principle of choice of methods
Variety of types of planning and forecasting allows you to choose the best way to solve a particular problem. Properly chosen methods significantly improve the quality of forecasting, as they ensure the completeness, reliability and accuracy of the forecast, as well as the opportunity to save time and reduce the cost of forecasting.
The choice of method is influenced by:
- the essence of the practical problem to be solved;
- dynamic characteristics of the forecasting object in the external environment;
- type and nature of available information, typical type of forecasting object;
- requirement regarding forecasting results and other specifics of the specific problem.
All these factors should be considered as a single system, while only insignificant factors can be excluded from consideration. On theIn practice, when choosing a forecasting method, it is recommended to consider two main factors - cost and accuracy.
When choosing a method, consider options:
- availability of statistical data for the required period;
- competence of forecaster, availability of equipment;
- required time to collect and analyze information.
Forecasts in various fields
The presented methods in this or that combination are used in various fields. Among the types of social forecasting, collective and individual intuitive methods can be distinguished. Mathematical methods are also widely used in this area. They are also the main type of economic forecasting. It is, in fact, a system of scientific research that has a quantitative and qualitative character. Used at the preliminary stage of developing economic solutions.
To make various types of forecasts, forecasting is often resorted to in such an area as sports. This applies to a wide variety of processes: the development of sports and its individual types, competitions, sports training systems, technical and tactical features, the emergence of new sports records, etc. Of the huge number of types of forecasting in sports, they use, in particular, scientific, empirical and intuitive methods: methods of logical analysis; expert assessments; extrapolation; analogies; modeling, etc.
Of particular interest is the compilation of forecasts in criminology, during which the future state of crime is known, factorsaffecting its changes, a criminological forecast is being developed. It allows you to establish the most general indicators that characterize the development (change) of crime in the future, to identify on this basis undesirable trends and patterns, to find ways to change them in the right direction.
There are several types of criminological forecasting: crime, the identity of the offender, factors and consequences of crime, measures to combat crime. There are also forecasting the development of the science of criminology, predicting crime and predicting individual criminal behavior.
The presented division of methods into groups is rather conditional. It should be noted that independent use of these groups of forecasting methods is impossible. Modern conditions (progress in science and technology, as well as the sophistication of connections in systems and their structure) necessitate the use of several forecasting methods to solve one problem. This led to the emergence of combined methods. Their use is especially relevant for complex socio-economic systems, when various combinations of forecasting methods can be used in the development of forecast indicators for each element of the system.
