As you know, the XXI century is called the century of information technology. Indeed, modern man operates with different methods of obtaining and processing information. Analytics plays an important role in the process of using information. What is analysis? What methods of evaluating information exist? Read on for answers to these and other questions.
What is analytics?
This word has Greek roots and literally translates as "the art of analysis". This term was used by Aristotle to refer to the technique of logical data mining.
Today, scientists give a broader interpretation of the concept. Analytics in the modern world is considered a part of logic (the art of reasoning), within which the doctrine of data analysis is considered. The operations of real or mental division of the whole (process, representation, relations between objects, etc.) into constituent elements are investigated.
What is data analysis? This concept is narrower than the term "analytics". Scientifically, data analysis is called the branch of computer science and mathematics, within which the construction and study of the mostgeneral computational algorithms and methods for extracting knowledge from information obtained experimentally. In other words, we are talking about a set of techniques related to information processing algorithms. Information analysis in the narrow sense is the process of studying, filtering, transforming (modeling) in order to extract useful data and make decisions.
Machine learning
It is considered today the most powerful and most common method of information analysis. Today, unfortunately, there are no machine learning algorithms that provide good processing of information of a more or less arbitrary nature. In this regard, specialists are forced to carry out a preliminary collection and processing of data in order to bring them into a form suitable for using the algorithm. As a rule, such processing is called featureselect or reprocessing. Most algorithms can use fixed length numbers.
At the same time, interest in algorithms based on neural networks has increased. The advantage is that they can be used not only for numbers, but also for objects that have additional (mostly geometric) properties. For example, you can analyze an image: the algorithm takes into account the value of the pixels, as well as their relative position. In a similar way, the initial data of an audio track or video sequence is evaluated.
Economic analysis as a science
Economic evaluation of data is a system of special knowledge based on the patterns of development and functioning of the economic complex, aimed atstudy of the methodology of analysis, diagnostics, planning and forecasting of financial and economic operations at the enterprise.
The subject of economic analysis is the economic activity of the organization, its socio-economic efficiency and final financial performance. The value of the latter is formed under the influence of subjective and objective factors. Indicators of financial and economic activity are reflected in the reporting system of the enterprise.
Purpose of information research
Evaluation of data in the economy provides the necessary number of parameters through which you can form an objective idea of the financial condition of the organization, its profits, losses, changes in the composition of liabilities and assets. With the help of analysis, you can determine the most rational and unprofitable areas of work, the distribution of financial, material and labor resources.
Dialectical method
This method of data evaluation involves the study of phenomena and processes in their dynamics, that is, in constant change. From this follows the main feature of the method - the need to compare certain indicators. You can compare values with different sources: the results of past years, planned indicators, achievements of competitors, etc.
According to the theory of materialistic dialectics, each phenomenon is seen as a unity and at the same time a struggle of opposites. From this follows the need to study internal contradictions, negative and positive aspectseach process.
When using the dialectical method of data evaluation, all interdependencies and relationships are taken into account. It is impossible to objectively analyze the process in isolation from other phenomena and events. The interdependence and interconnection of economic operations necessitates the use of complex methods for analyzing economic activity. Only a comprehensive study of information allows you to correctly assess the results of the work, reveal the reserves.
Deduction and induction
There is a causal relationship between many processes and events. It means that one thing follows from another. Establishing a causal relationship is the most important task in the economic evaluation of data. As a result, the analysis is more accurate and objective. This, in turn, allows us to quantify the data, determine the degree of influence of certain factors on the operation of the enterprise.
Induction involves the study of processes from particular to general: from factors to conclusions, from causes to results. Deduction is an inverse method that involves research from the general to the particular. In this case, a kind of "dismemberment" of the phenomenon into elements takes place.
Systemacity
When using the dialectical approach to data evaluation, each phenomenon, process, event must be considered as a set of many components that are closely related to each other. Maximum detailing is carried out during the implementation of a systematic approach. When describing data types,their characteristics, determination of the degree of influence of factors on them, etc., the most important, important thing in the object under study is revealed. A systematic approach allows you to build an approximate scheme of the process, establish its key components, their subordination, functions and, as a result, reveal the logical and methodological model of the analysis.
In the economic assessment, after examining certain aspects of the organization's activities, their interdependence, subordination, the collected data are summarized. At the same time, key and determining ones are singled out from the entire amount of data and factors. The results of economic activity mainly depend on them.
Economic models
For the systematic classification of data, their evaluation and processing, it is necessary to build a scheme corresponding to the tasks and ultimate goals of the study. Depending on the object under study, optimization and equilibrium models are distinguished. The former are used to describe the behavior of economic entities that achieve their goals with the available opportunities. Equilibrium models are used to determine the result of the interaction of a group of subjects, to identify the conditions for the compatibility of their tasks and goals.
Analysis methods
The results of the interaction of economic entities will depend on the period of time within which their behavior is being studied. Accordingly, methods of comparative statistics, statistical and dynamic analysis are distinguished.
The first is to compare the results of the statistical evaluation of activities in different time periods. Dynamic analysis is used to determine the naturechanges in economic indicators between given points in time and determining the factors that determine these changes. Statistical evaluation involves the study of actions at a certain moment. For example, you can determine how, given supply and demand, the cost of a product is formed.
The methodology of macroeconomic assessment is based on the intersection of three areas of knowledge: mathematics, statistics and economics. Economic methods are: comparison, grouping, graphical and balance analysis.
Mathematical techniques are divided into 3 groups:
- Economic. These include matrix methods, theories of input-output balance, production functions.
- Optimal programming techniques (nonlinear, linear, dynamic) and economic cybernetics.
- Methods for studying the decision-making process and transactions. This group contains theories of queuing, games, graphs.
Comparative analysis
Comparison is a comparison of researched data and facts. In practice are used:
- Horizontal analysis. It is necessary to identify relative and absolute deviations of the actual value of indicators from the baseline.
- Vertical analysis. It is used to study the structure of phenomena.
- Trend analysis. It is used to study the relative growth rates of indicators over several years compared to the level of the base year.
Balance analysis
It lies incomparative measurement of two sets of indicators tending to equilibrium. As a result, the researcher determines a new - balancing - indicator.
For example, when assessing the degree of provision of an enterprise with raw materials, they compare the needs for it, the sources for covering these needs and set a balancing indicator - an excess or shortage of materials.
As an auxiliary balance method, it is used when checking the result of calculating the influence of factors on the total performance indicator. If the sum of the impact is equal to the deviation from the base value, then the calculations are correct.
Extra
Graphs are used to scale indicators. Values and their dependence is described by constructing geometric shapes. It must be said that the graphical method in the analysis of independent significance does not matter. It is only used to illustrate the changes.
Index evaluation is based on relative values, which expresses the ratio of the level of the phenomenon under consideration to the base level. Several types of indices are used in statistics: harmonic, arithmetic, aggregate, etc.
If you use index recalculations and build a time series that reflects, for example, the release of goods in value terms, you can objectively assess the dynamics.
Regression (stochastic) and correlation methods are used to determine the level of connection between parameters that are functionally independent of each other. Through correlationyou can:
- Make a model of existing factors.
- Quantify the strength of the connection.
Analysis in sociology
Description of any phenomena can be carried out in a variety of ways. One of the most common methods of analysis in sociology is observation. During it, you can quantify the data through:
- Psychological scaling. Typically, scores are used to summarize observations.
- Measuring time (timekeeping).
Another approach is the time sampling method. When using it, certain time periods are selected from a single process under study to consolidate information. They are considered representative of a longer period. In real research, quantitative and qualitative descriptions of phenomena are usually carried out in combination.
Quantitative indicators can be recorded during the observation or generalized after its completion, included in a retrospective report. The general impressions of the researcher serve as the basis for retrospective evaluation. For long-term follow-up, they may, for example, include the frequency of any of the episodes being studied. Quantitative indicators can thus be included in value judgments. For example, "he rarely goes to school", "she always forgets her textbook", etc.
In addition to the evaluative description of events, the researcher can use a point assessment of his impressions. These figures reflectcharacteristic of long-term uncontrolled observations in everyday life. As some studies show, they may well be used as one of the main or only criteria for the adequacy of psychological tests or characteristics of an individual.
