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. A special place in the process of using information is occupied by analytics. What is analysis? What methods of information assessment exist? Read the 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 part of the logic (the art of reasoning), in which the doctrine of data analysis is considered. The operations of real or mental partitioning of the whole (process, presentation, relationships between objects, etc.) into constituent elements are investigated.
What is data analysis? This concept is narrower in relation to the term "analytics". Scientifically, data analysis is the branch of informatics and mathematics, in the framework of which the construction and study of the most common computational algorithms and methods for extracting knowledge from information obtained experimentally is carried out. In other words, we are talking about a set of techniques associated with information processing algorithms. The analysis of information in the narrow sense refers to 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 analyzing information. 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 preliminary data collection and processing in order to bring them into a form suitable for using the algorithm. As a rule, such processing is called feature-selection or reprocessing. Most algorithms can use fixed-length numbers.
At the same time, interest in algorithms based on neural networks has now increased. The advantage is that they can be used not only for numbers, but also for objects that have additional (mainly geometric) properties. For example, you can analyze the image: the algorithm takes into account the value of the pixels, as well as their relative position. Similarly, the source data of an audio track or video is evaluated.
Economic analysis as a science
The economic assessment of data is a system of special knowledge based on the laws of development and functioning of the economic complex, aimed at studying the methodology of analysis, diagnosis, 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 indicators. The significance of the latter is formed under the influence of subjective and objective factors. Indicators of financial and economic activity are reflected in the reporting documentation system of the enterprise.
Information Research Purpose
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. Using the analysis, you can determine the most rational and non-profitable areas of work, the distribution of financial, material and labor resources.
Dialectical method
This method of evaluating data involves the study of phenomena and processes in their dynamics, i.e., 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: results of past years, planned indicators, achievements of competitors, etc.
According to the theory of materialist dialectics, each phenomenon is considered as a unity and at the same time the struggle of opposites. This implies the need to study the internal contradictions, negative and positive aspects of each process.
When using the dialectical method for evaluating data, all interdependencies and relationships are taken into account. It is impossible to objectively analyze the process apart from other phenomena and events. The interdependence and interconnection of economic operations necessitates the use of complex methods of analysis of economic activity. Only a comprehensive study of the information allows you to correctly evaluate the results of work, to reveal reserves.
Deduction and induction
There is a causal relationship between many processes and events. It means that one phenomenon arises from another. Establishing a causal relationship is the most important task of economic data assessment. As a result, the analysis is more accurate and objective. This, in turn, allows you to quantify the data, to determine the degree of influence of certain factors on the work of the enterprise.
Induction involves the study of processes from particular to general: from factors to conclusions, from reasons to results. Deduction is the reverse method, involving research from the general to the particular. In this case, there is a kind of "dissection" of the phenomenon into elements.
Systematic
When using the dialectical approach to data evaluation, each phenomenon, process, event must be considered as a complex 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, determining the degree of influence of factors on them, etc., the most important, important in the studied object is revealed. The systematic approach allows you to build an approximate diagram of the process, establish its key components, their subordination, functions and, as a result, reveal the logical and methodological model of analysis.
In an economic assessment, after researching individual aspects of an organization’s activities, their interdependence, and subordination, the collected data are generalized. At the same time, key and determining ones are distinguished from the total amount of data and factors. The results of economic activity mainly depend on them.
Economic models
For a systematic classification of data, their evaluation and processing, it is necessary to build a scheme that meets the objectives and ultimate goals of the study. Depending on the object being studied, 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 compatibility conditions of their tasks and goals.
Analysis methods
The results of the interaction of business entities will depend on the period of time within which their behavior is studied. Accordingly, methods of comparative statistics, statistical and dynamic analysis are distinguished.
The first is reduced to comparing the results of a statistical assessment of activity at different time intervals. Dynamic analysis is used to determine the nature of changes in economic indicators between given points in time and determine the factors that determine these changes. Statistical evaluation involves the study of actions at a certain point. For example, you can determine how, for given supply and demand, the value of the product is formed.
The macroeconomic assessment methodology is based on the intersection of three areas of knowledge: mathematics, statistics and economics. Economic methods are: comparison, grouping, graphic and balance sheet analysis.
Mathematical techniques are divided into 3 groups:
- Economic. These include matrix methods, theories of intersectoral balance, production functions.
- Techniques for optimal programming (non-linear, linear, dynamic) and economic cybernetics.
- Methods of studying the decision-making process and transactions. In this group there are theories of queuing, games, graphs.
Comparative analysis
Comparison is a comparison of the investigated 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 base.
- 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 in comparison with the level of the base year.
Balance sheet analysis
It consists in a comparative measurement of two sets of indicators striving for 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 of covering these needs and establish a balancing indicator - excess or deficit of materials.
As an auxiliary, the balance method is used when checking the result of calculating the influence of factors on the aggregate effective indicator. If the sum of the impact is equal to the deviation from the base value, then the calculations are correct.
Additionally
For a large-scale image indicators used graphs. Values and their dependence are described by constructing geometric shapes. I must say that the graphical method in the analysis does not have independent significance. It is used only to illustrate the changes.
The index score is based on relative values, through which the ratio of the level of the phenomenon under consideration to the base level is expressed. In statistics, several types of indices are used: harmonic, arithmetic, aggregate, etc.
If you use index counts and build a time series that reflects, for example, the release of goods in value terms, you can objectively evaluate the dynamics.
Regression (stochastic) and correlation methods are used to determine the level of communication between parameters that are functionally independent of each other. By correlation, you can:
- Make a model of existing factors.
- Quantify the tightness of communication.
Analysis in sociology
The description of any phenomena can be carried out in a variety of ways. One of the common methods of analysis in sociology is observation. During it, you can get a quantitative assessment of the data by:
- Psychological scaling. As a rule, points are used to summarize the observations.
- Time measurement (timing).
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 for a longer period. In real research, quantitative and qualitative descriptions of phenomena are usually carried out in a complex.
Quantitative indicators can be recorded during the observation or summarized after its completion, included in a retrospective report. The basis for the retrospective assessment is the general impression of the researcher. With prolonged observation, they may, for example, include the frequency of any episodes studied. Quantitative indicators can thus be included in value judgments. For example, "he rarely goes to school," "she always forgets the textbook," etc.
In addition to evaluating the events, the researcher can use a point estimate of his impressions. Such indicators reflect the characteristic of prolonged uncontrolled observations in everyday life. As some studies show, they may well be used as one of the main or the only criterion for the adequacy of psychological tests or characteristics of an individual.