Search

The paper considers missing data management task in smart systems. The main strategies of missing data management in handling missing data are analyzed. A prediction method for probable recovery of partially missing or completely lost data based on the improvement of an ensemble of two GRNNs by the additional use of extended-input SGTM neural-like structure is proposed. The latter is used to increase the accuracy of the procedure of weighted summation with the displacement of the outputs of both GRNN networks in comparison with existing methods. The flowchart of the ensemble is given. The training algorithm is presented, as well as the detailed procedure of its use. The improved ensemble prediction method has been tested for the task of completing the gaps in the real dataset reflecting air condition monitoring. The optimal parameters of the component operation of the improved ensemble have been determined experimentally. The efficiency of its performance has been evaluated by its experimental comparison with methods of the same class based on the MAPE and RMSE accuracy indicators. The minimum value of an application error of the developed method for solving the task in comparison with existing ones for the specified accuracy indicators is established. The time delays in applying the methods examined have been determined experimentally. The limitations of the improved method and the prospects for further research are described.

A significant influence on the choice of both specialty and specific educational institution made by each entrant during the admission campaign is the assessment of the applicant’s chances of admission. One possible way to solve this task is to build specialized web services using machine learning methods based on the processing of entry data for previous years. In this paper, research and experimental analysis of existing machine learning methods to solve this task were made. Experimental evaluation of Linear Regression, Stochastic Gradient Descent (SGD), Tree, AdaBoost, Random Forest, Support Vector Machine (SVM) and k-nearest neighbors’ algorithm (k-NN) was performed on a real data set. It was found that the highest prediction accuracy based on all accuracy indicators was obtained using methods based on Linear Regression and SGD. Since the latter method is the basis of a number of algorithms for training artificial neural networks, in the future it is advisable to evaluate and improve existing and develops, ANN-based approaches to improve the accuracy of the task in this study.

An effective solution of the prediction tasks requires the high accuracy of the result with minimal resource and time costs for the operation of the chosen algorithm. In cases when high accuracy of the received result has the first priority, it is expedient to use ensemble learning. This paper describes a prediction method using a new, stacking-based GRNN ensemble model. Each member of the developed ensemble processes its own dataset, where the vectors of the original set of data are randomly shifted relative to the current point. The authors chose SGTM neural-like structure as a meta-algorithm for the formation of the result of the ensemble. This choice is argued by the high accuracy and speed of its work. The results of a number of experimental studies on the optimal parameters selection of the developed ensemble are described. A comparison of the efficiency of its work with a number of known predictors was done. Prospects for further research are described.

Today, the fast development of the hardware for the Internet of things systems creates conditions for the development of IoT based Services of various purposes. The imperfect systems of collecting, aggregation and the transmission of large volumes of various types of data, fixed by sensors of IoT devices, as well as possible failures of the latter, cause the occurrence of missing data problems. The paper proposes a regression approach to solving the task of missed data recovery. The authors have developed a composition of the method of the missing data recovery for IoT systems based on the use of the Ito decomposition and the AdaBoost algorithm. We transform each data vectors by using Ito decomposition, and searching the coefficients of this decomposition scheme using AdaBoost algorithm. Increasing the dimensionality of the input space due to the use of the second-degree Ito decomposition scheme, as well as its high approximation properties, allowed to increase the accuracy of filling the missed values by the AdaBoost regressor at more than 6% (MAPE). It has been established that the developed method provides the highest accuracy of filling missed data based on all other indicators (MAE, RMSE, SMAPE) among the considered regression methods.