卷 20, 编号 3 (2021)
Artificial intelligence, knowledge and data engineering
Analytical Review of Automatic Systems for Depression Detection by Speech
摘要
In recent years the interest in automatic depression detection has grown within medical and scientific-technical communities. Depression is one of the most widespread mental illnesses that affects human life. In this review we present and analyze the latest researches devoted to depression detection. Basic notions related to the definition of depression were specified, the review includes both unimodal and multimodal corpora containing records of informants diagnosed with depression and control groups of non-depressed people. Theoretical and practical researches which present automated systems for depression detection were reviewed. The last ones include unimodal as well as multimodal systems. A part of reviewed systems addresses the challenge of regressive classification predicting the degree of depression severity (non-depressed, mild, moderate and severe), and another part solves a problem of binary classification predicting the presence of depression (if a person is depressed or not). An original classification of methods for computing of informative features for three communicative modalities (audio, video, text information) is presented. New methods for depression detection in every modality and all modalities in total are defined. The most popular methods for depression detection in reviewed studies are neural networks. The survey has shown that the main features of depression are psychomotor retardation that affects all communicative modalities and strong correlation with affective values of valency, activation and domination, also there has been observed an inverse correlation between depression and aggression. Discovered correlations confirm interrelation of affective disorders and human emotional states. The trend observed in many reviewed papers is that combining modalities improves the results of depression detection systems.
Informatics and Automation. 2021;20(3):497-529
497-529
Model OF Information Interaction between Elements of Multilevel System of Digital Twins
摘要
One of the solutions to the problem of spatio-temporal data anisotropy is the use of a multilevel system of digital twins based on the corresponding industry models and the updated archive data base. The application of this approach has successfully proved itself in information systems for monitoring the parameters of the geomagnetic field and its variations, providing spatio-temporal interpolation of geomagnetic data with an accuracy of 0.81 nT in magnetically quiet periods. At the same time, the problem of information interaction between the levels of the system of digital twins remained unresolved, which is greatly aggravated by the constantly growing volume of data and their heterogeneous nature. The paper proposes a solution to the indicated problem by means of a formalized mechanism for packaging space-time information, in which the identification of data sources is performed on the basis of a hierarchical binary tokenization system. In addition, the proposed software implementation of such an approach is considered, a distinctive feature of which is the combination of traditional clientserver and innovative serverless architectures to implement a highly loaded reactive web application for working with analyzed data. The main stages of the implementation of information interaction are highlighted and programmatically formalized - from obtaining initial information from its sources to verifying data, analyzing them, processing and forming the output information flow of the system. The results of the computational experiments carried out on the example of the problem of monitoring the parameters of the Earth's magnetic field and its variations confirmed the effectiveness of the proposed solutions, expressed both in increasing the reactivity of web-based applications and in increasing the computational speed of formation and filling of information storages that aggregate information from distributed heterogeneous sources.
Informatics and Automation. 2021;20(3):530-561
530-561
Formalization of Regulatory Texts
摘要
Modern information technologies provide text manipulation processes with high efficiency. First of all, this means storing, editing, and formatting texts and their components. Having achieved significant success in developing tools for content-free computer text processing, researchers faced problems with their content processing. Therefore, further steps in this direction are associated with the creation, among other things, of methods for automated purposeful manipulation of texts, taking into account their content. The analysis of works devoted to the study of the problems of formal presentation of texts and their subsequent use is carried out. Despite a number of successful projects, the challenges of solving the problem of the relationship between the content of the text and its meaning remain relevant. It seems that formalization of a General-purpose text while preserving its semantics is not feasible at this stage. However, there are types of texts that can be formalized while preserving their semantics. One of them is a regulatory text type, which is essentially a verbally expressed algorithm for a sequence of targeted actions. It is distinguished by logic and accuracy (lack of allegories), coherence and integrity, clarity, understandability (due to the lack of emotional coloring and figurative means), accessibility (due to the use of specific terminology). In other words, when developing regulatory texts, they usually try to display the mechanisms of the described actions as clearly as possible. Purpose: development of a method for formalizing a regulatory text while preserving its semantics. Methods: structural linguistics, representation of objects in the form of an ontology, constructive algorithms. The use of this method is demonstrated by describing the solution of a system of algebraic equations. Results: method for constructing a mathematical model of a regulatory text. Practical relevance: the application of the developed method makes it possible to develop software systems for building libraries of individual subject areas, develop tools for evaluating regulatory texts for their certainty, completeness, connectivity and other characteristics, as well as simulators and self-learning tools.
Informatics and Automation. 2021;20(3):562-590
562-590
Behavioral Functions Implementation on Spiking Neural Networks
摘要
The question of behavioral functions modeling of animals (in particular, the modeling and implementation of the conditioned reflex) is considered. The analysis of the current state of neural networks with the possibility of structural reconfiguration is carried out. The modeling is carried out by means of neural networks, which are built on the basis of a compartmental spiking model of a neuron with the possibility of structural adaptation to the input pulse pattern. The compartmental spike model of a neuron is able to change its structure (the size of the cell body, the number and length of dendrites, the number of synapses) depending on the incoming pulse pattern at its inputs. A brief description of the compartmental spiking model of a neuron is given, and its main features are noted in terms of the possibility of its structural reconfiguration. The method of structural adaptation of the compartmental spiking model of the neuron to the input pulse pattern is described. To study the work of the proposed model of a neuron in a network, the choice of a conditioned reflex as a special case of the formation of associative connections is justified as an example. The structural scheme and algorithm of formation of a conditioned reflex with both positive and negative reinforcement are described. The article presents a step-by-step description of experiments on the associative connection’s formation in general and conditioned reflex (both with positive and negative reinforcement), in particular. The conclusion is made about the prospects of using spiking compartmental models of neurons to improve the efficiency of the implementation of behavioral functions in neuromorphic control systems. Further promising directions for the development of neuromorphic systems based on spiking compartmental models of the neuron are considered.
Informatics and Automation. 2021;20(3):591-622
591-622
Efficient natural language classification algorithm for detecting duplicate unsupervised features
摘要
This paper focuses on capturing the meaning of Natural Language Understanding (NLU) text features to detect the duplicate unsupervised features. The NLU features are compared with lexical approaches to prove the suitable classification technique. The transfer-learning approach is utilized to train the extraction of features on the Semantic Textual Similarity (STS) task. All features are evaluated with two types of datasets that belong to Bosch bug and Wikipedia article reports. This study aims to structure the recent research efforts by comparing NLU concepts for featuring semantics of text and applying it to IR. The main contribution of this paper is a comparative study of semantic similarity measurements. The experimental results demonstrate the Term Frequency–Inverse Document Frequency (TF-IDF) feature results on both datasets with reasonable vocabulary size. It indicates that the Bidirectional Long Short Term Memory (BiLSTM) can learn the structure of a sentence to improve the classification.
Informatics and Automation. 2021;20(3):623-653
623-653
Mathematical modeling and applied mathematics
Construction of Complex Schedules for Execution of Task Packages at Forming Sets in Specified Directive Terms
摘要
The current state with the solution of the problem complex planning of the execution of task packets in multistage system is characterized by the absence of universal methods of forming decisions on the composition of packets, the presence of restrictions on the dimension of the problem and the impossibility of guaranteed obtaining effective solutions for various values of its input parameters, as well the impossibility of registration the condition of the formation of sets from the results. The solution of the task of planning the execution of task packets in multistage systems with the formation of sets of results within the specified deadlines has been realized of authors in article. To solve the planning problem, the generalized function of the system was decomposed into a set of hierarchically interrelated subfunctions. The use of decomposition made it possible to use a hierarchical approach for planning the execution of task packets in multistage systems, which involves defining solutions based on the composition of packets at the top level of the hierarchy and scheduling the execution of packages at the bottom level of the hierarchy. The theory of hierarchical games is used to optimize solutions for the compositions of task packets and schedules for their execution is built, which is a system of criteria at the decision-making levels. Evaluation of the effectiveness of decisions by the composition of packets at the top level of the hierarchy is ensured by the distribution of the results of task execution by packets in accordance with the formed schedule. To evaluate the effectiveness of decisions on the composition of packets, method for ordering the identifiers of the types of sets with registration of the deadlines and a method for distributing the results of the tasks performed by packets has been formulated, which calculates the moments of completion of the formation of sets and delays with their formation relative to the specified deadlines. The studies of planning the process of the executing task packages in multistage systems have been carried out, provided that the sets are formed within specified deadlines. On their basis, conclusions, regarding the dependence of the planning efficiency from the input parameters of the problem, were formulated.
Informatics and Automation. 2021;20(3):654-689
654-689
Method for Optimizing of Mobile Robot Trajectory in Repeller Sources Field
摘要
The article discusses the procedure for correcting the trajectory of a robotic platform (RTP) on a plane in order to reduce the probability of its defeat/detection in the field of a finite number of repeller sources. Each of these sources is described by a mathematical model of some factor of counteraction to the RTP. This procedure is based, on the one hand, on the concept of a characteristic probability function of a system of repeller sources, which allows us to assess the degree of influence of these sources on the moving RTP. From this concept follows the probability of its successful completion used here as a criterion for optimizing the target trajectory. On the other hand, this procedure is based on solving local optimization problems that allow you to correct individual sections of the initial trajectory, taking into account the location of specific repeller sources with specified parameters in their vicinity. Each of these sources is characterized by the potential, frequency of impact, radius of action, and parameters of the field decay. The trajectory is adjusted iteratively and takes into account the target value of the probability of passing. The main restriction on the variation of the original trajectory is the maximum allowable deviation of the changed trajectory from the original one. If there is no such restriction, then the task may lose its meaning, because then you can select an area that covers all obstacles and sources, and bypass it around the perimeter. Therefore, we search for a local extremum that corresponds to an acceptable curve in the sense of the specified restriction. The iterative procedure proposed in this paper allows us to search for the corresponding local maxima of the probability of RTP passage in the field of several randomly located and oriented sources, in some neighborhood of the initial trajectory. First, the problem of trajectory optimization is set and solved under the condition of movement in the field of single source with the scope in the form of a circular sector, then the result is extended to the case of several similar sources. The main problem of the study is the choice of the General form of the functional at each point of the initial curve, as well as its adjustment coefficients. It is shown that the selection of these coefficients is an adaptive procedure, the input variables of which are characteristic geometric values describing the current trajectory in the source field. Standard median smoothing procedures are used to eliminate oscillations that occur as a result of the locality of the proposed procedure. The simulation results show the high efficiency of the proposed procedure for correcting the previously planned trajectory.
Informatics and Automation. 2021;20(3):690-726
690-726
Probabilistic Analysis of Generalised Statistic Model for Multipath Channel of Simo Sistems with Fading and Correlated Shadowing
摘要
The paper considers the problem of analysis of the information transmission process by multi-element communication systems in presence of a multipath signal propagation channel. To generalize the propagation effects, the model of the κ–μ fading channel with correlated shadowing was assumed, and the technology used for organizing a multi-element system was the SIMO system, equipped with the maximum-ration combiner of the signal on the receiving side. To describe the characteristics of the information transfer process, an approach based on the higher-order statistics of the ergodic capacity was used. Closed-form analytical expressions for arbitrary-order capacity higher-order statistics were obtained for the channel model under consideration. The behavior of the first four statistics (ergodic capacity, its reliability, skewness and kurtosis coefficients) is analyzed depending on the channel parameters (the number of multipath propagation clusters, the ratio of power of the dominant components to the total power of multipath waves, the degree of shadowing of the dominant components, and the shadowing correlation coefficient). Within the framework of the study, 4 distinct situations of the assumed channel model behavior were considered, which significantly differ in their properties. It is noted that, in contrast to the capacity, its higher-order statistics are significantly more sensitive to the channel parameters and, as a result, are more significant indicators of fluctuations in the information transfer rate within the communication channel. The existence of a pronounced extremum (minimum) of the reliability ergodic capacity dependence from the signal-to-noise ratio was established. It should be accounted for in practical applications, when the requirements of the signal-to-noise ratio that guarantees the desired communication link quality are set.
Informatics and Automation. 2021;20(3):727-749
727-749