Ўзбекистон республикаси ахборот технологиялари ва коммуникацияларини ривожлантириш вазирлиги муҳаммад ал-хоразмий номидаги

 
COMPARATIVE CHARACTERISTICS OF THE BASIC METHODS IN 
INTELLECTUAL DATA ANALYSIS 
S. Makhmudjanov (assistant teacher of TUIT) 
S. Mamajonov (master of TUIT) 
A characteristic feature of Data Mining is the active use of classification, 
clustering and forecasting methods used to identify implicit patterns and properties 
present in the data. Let us consider, without mathematical detailing, some applied 
examples of the application of these methods in solving practical problems. 
Along with the search for the most general types of patterns that may be present 
in the data, groups of more specific, particular tasks of data analysis are also 
distinguished. Despite the extensive scope of Data Mining application in business, 
medicine or government, the vast majority of these tasks can be combined into a 
relatively small number of groups. 
There are a large number of different grounds for stratification, categorization, 
classification of a significant number of existing and newly developed Data Mining 
methods. For example, you can find classifications according to the principle of work 
with the original training data (whether they undergo changes or not as a result of 
processing), according to the type of result obtained. , according to the types of 
mathematical apparatus used (statistical and cybernetic), etc. For example, according 
to the type of mathematical apparatus used, as a rule, the following main groups of 
methods Data Mining: 

---

154 
1. Descriptive analysis and description of the initial data, preliminary analysis 
of the nature of statistical data (testing hypotheses of stationarity, normality, 
independence, homogeneity, assessment of the form of the distribution function, its 
parameters, etc.). 
2. Multivariate statistical analysis (linear and nonlinear discriminant analysis, 
cluster analysis, component analysis, factor analysis, etc.). 
At the end of various approaches to the classification of Data Mining methods, 
we will give an example of a comparative analysis of the most widely used methods 
among themselves, using the following rating scale as a characteristic of each of the 
attributes: “extremely low, very low, low / neutral, neutral / low, neutral, neutral / 
high, high, very high ”(Table 1). It can be seen that none of the methods can be 
recognized as the only effective one, having an obvious superiority over other 
methods. 
The scope of IAD tools is not limited exclusively to business areas, the main 
indicator of efficiency in which is profit. Obviously, such a toolkit can and finds 
application in other areas of human activity, the functioning of which is accompanied 
by the generation and analysis of various data. One of the most important areas in 
which IAD methods are being actively adapted is medicine. 
Medicine. For example, these methods were used to create diagnostic and 
predictive algorithms in oncology, neurology, pediatrics, psychiatry, gynecology and 
other fields. Based on the results obtained, expert systems were built for making 
diagnoses using rules describing combinations of symptoms of different diseases. The 
rules help to choose the indications (contraindications), predict the outcomes of the 
prescribed course of treatment. In molecular genetics and genetic engineering, this is 
the definition of markers, which are understood as genetic codes that control those or 
other phenotypic traits of a living organism. There are several large firms specializing 
in the use of IAD for decoding the human and plant genomes. In applied chemistry, 
these methods are used to clarify the structural features of chemical compounds. 
An example of a comparative analysis of Data Mining methods. Table 1.