<abstract>
<p>The paper considers the problem of handling short sets of medical data. Effectively solving this problem will provide the ability to solve numerous classification and regression tasks in case of limited data in health decision support systems. Many similar tasks arise in various fields of medicine. The authors improved the regression method of data analysis based on artificial neural networks by introducing additional elements into the formula for calculating the output signal of the existing RBF-based input-doubling method. This improvement provides averaging of the result, which is typical for ensemble methods, and allows compensating for the errors of different signs of the predicted values. These two advantages make it possible to significantly increase the accuracy of the methods of this class. It should be noted that the duration of the training algorithm of the advanced method remains the same as for existing method. Experimental modeling was performed using a real short medical data. The regression task in rheumatology was solved based on only 77 observations. The optimal parameters of the method, which provide the highest prediction accuracy based on MAE and RMSE, were selected experimentally. A comparison of its efficiency with other methods of this class has been performed. The highest accuracy of the proposed RBF-based additive input-doubling method among the considered ones is established. The method can be modified by using other nonlinear artificial intelligence tools to implement its training and application algorithms and such methods can be applied in various fields of medicine.</p>
</abstract>
Background:
Today, using of systems on the base of Internet of Things (ІоТ) devices is very widespread in
various applications. Intellectual analysis of the data collected by similar devices is an important task for efficient and
successful functioning of such systems. In particular, the reliability of such kind of analysis has greatly influence on the
ability to partially or fully automate certain processes or subsystems. However, imperfect devices of data collection,
transportation errors, etc. cause data missing to appear. A number of limitations cause this problem, and in the work, they
makes it impossible an effective intellectual analysis for specific use. That is why the scientific and applied problem of
effectively filling the missing in the data collected by the sensors of specific characteristics should be considered.
Methods:
The authors propose a new prediction method for solving this problem based on the use of General Regression
Neural Networks (GRNN).
Results:
The possibility of approximation and partial elimination of the error of computational intelligence of this type has
been analytically proved. A cascade of two sequentially connected GRNN was developed. The optimal parameters of the
developed cascade were selected. The simulation of its work was performed to solve the problem of recover missing sensor
data in the dataset for monitoring the state of air environment. A high number of missing for one reason or another
characterizes this real data set, collected by IoT device.
Conclusion:
High accuracy of cascade operation in comparison with existing methods of this class is inserted. All
advantages and disadvantages are described. Perspectives of further research are outlined.
Neuro-fuzzy models of management nowadays are becoming more widespread in various industries. Many papers deal with the synthesis of neuro-fuzzy models of diagnostics in economics, medicine, and industrial tasks. Most of them are based on iterative topologies of artificial neural networks and traditional fuzzy inference systems. The latter ones do not always ensure high accuracy, which affects the entire system that is being developed. This paper presents a new neuro-fuzzy diagnostic system based on non-iterative ANN and a new fuzzy model, a T-controller. The flowchart of the system proposed is given. All the operation stages of the system are described in detail, from the collection and preliminary processing of data to two different stages of diagnostics. The last stage can be performed manually or using a T-controller. Simulation of the system is conducted by means of a real set of data. The task was to predict the generator power based on a set of 13 independent variables. To improve the accuracy of the system performance, the Padé polynomial has been used, the coefficients of which are synthesized on the basis of a pre-trained SGTM neural-like structure using an optimization simulated annealing simulation. High accuracy of the system performance is shown using various performance indicators. The coefficients of the required polynomial are synthesized, which allows of fulfilling the task in a more understandable form.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.