Now in the world there are a lot of diseases threatening humanity and for these medical treatments and diagnostic methods have also improved both in practical and statistical. This paper discusses one of the most leading diseases, myocardial infarction. The mortality results showed that 18.8% of the patients hospitalized with MI diagnosis had died. Gender results revealed that around 28% of the deceased MI patients were women and 14% of them were men. So, by these statistics there are vital needs emerg ing in the diagnostician and prediction for making disease prevention and mitigation plans. To do that, one of cutting-edge methods is classical Ml prediction algorithms. The experiments and results showed that GUSTO-I trial data was used to train models in different kinds of ML algorithms, in terms of accuracy, the Random Forest algorithm showed slightly better results. And, to evaluate the model different methods are used including cross-validation and metrics are precision, Recall (Sensitivity), F1 Score, Con fusion Matrix but the final accuracy metric was Confusion Matrix. In the end, to predict long term mortality from myocardial infarction, a machine learning model trained and analyzed that for the GUSTO-I cardio data, the slightly effective algorithm is random forest.
| Mualliflar | Sharipov, D.K., Saidov, A.D., Шарипов, Д.К., Саидов, А.Д. |
|---|---|
| Jurnal | Ҳисоблаш ва амалий математика муаммолари |
| Nashr sanasi | 2024-10-11 |
| Son | 4 |
| Betlar | 135-147 |
| Til | Ingliz |
ml, logistic regression, decision trees, random forest, GUSTO-I, myocardial infarction, cross-validation, машинное обучение, логистическая регрессия, дерево решений, случайный лес, GUSTO-I, инфаркт миокарда, кросс-валидация
The article discusses modeling the quality management system of educational information systems (EIS) using fuzzy cognitive maps, as one of the most effective approaches to solving weakly formalized problems. To…
The article analyzes the work done so far to create a high-quality computer translation. It also describes the approaches, challenges, and solutions in achieving semantically accurate translation in computer…
Regimes of stochasticity and chaos are allowed in nonlinear reaction-diffusion systems that describe chemical-reactive, astrophysical, thermal, thermonuclear, plasma, epidemiological and demographic processes. The work…
The article considers a solution to a two-class recognition problem using metric al gorithms based on k-nearest neighbors. The proposed solution options are: the classical KNN algorithm based on a nonparametric estimate…
In the modeling of a number of practical problems, it is brought to the mixed problems imposed on the system of quasi-linear hyperbolic equations. Since there is no exact solution to such problems, they are solved using…
The paper proposes a method that belongs to the group of intensional methods of pattern recognition, the essence of which is the transition from a system of initial features to a reduced system of models of…
In this work, we consider system of linear and nonlinear Volterra integral equations (VIEs) of the first kind. At first, we have converted system of VIEs of the first kind to an equation of the second kind by…
A numerical study was conducted based on the proposed mathematical apparatus of non-stationary technology for filtering a low-concentration solution through a porous medium, taking into account the process of…
The construction of an optimal interpolation formula within the Hilbert space is the focus of this paper. The difference between a given function and an interpolation formula is estimated by the norm of the error…
In this article, within the framework of a mathematical model of three-phase filtration of incompressible and immiscible liquids and gas without taking into account the difference between phase pressures and…