Subject of research. Evaluation of the impact of priority data transmission in multi-channel systems of high load, presented as a queuing system with prioritization of applications. Method. The impact of priorities on...
详细信息
This paper overviews the Smart Home User Interfaces (UI) in the following aspects: general definition of Human-Machine Interface (HMI), types of common Smart Home UI, market analysis of HMI, voice assistants and chatb...
详细信息
The active development of the Internet of Things (IoT) in recent years has increased people's need for control smart devices for home. At the same time, the complexity of these devices is growing. Therefore, Smart...
详细信息
We suggest a new quantum-like approach to study distributed intelligence systems (DIS) consisting of natural (owners) and artificial (avatars) intelligence agents organized in a scale-free network. We demonstrate the ...
详细信息
This paper addresses the problem of deploying complex systems in Kubernetes clusters. It discusses using the OperatorSDK framework supported by RedHat as a basis for implementing the Kubernetes operator for Lightweigh...
详细信息
In this paper the problem of testing decision making systems for MEC platforms was formulated. Methods and means of organizing the introduction of network delays as part of the emulation system of MEC platforms LWMECP...
详细信息
In this paper, we investigate the effectiveness of various machine learning methods for the multimodal classification of personality traits using the HEXACO model and open data from social media users. Particular atte...
详细信息
ISBN:
(数字)9798331511241
ISBN:
(纸本)9798331511258
In this paper, we investigate the effectiveness of various machine learning methods for the multimodal classification of personality traits using the HEXACO model and open data from social media users. Particular attention is paid to the analysis of data obtained from user profiles, including statistical, visual, and audio characteristics. For each personality trait from the HEXACO model, a machine learning model is trained that determines the manifestation degree of this trait (low, medium, high). The influence of different types of information, both individually and in combination, on the weighted F1-score is considered. The study provides a comparative analysis of traditional machine learning methods that use all features and approaches with the selection of the most significant features. A strategy for combining data from different modalities using early merger is evaluated. The experimental results demonstrate that early merger models using selected features from different modalities provide the best result of the weighted F1-score, which varies depending on the personality trait from 0.76 to 0.85.
The following algorithms for constructing predictive models of key quality indicators of polymer film materials are considered and implemented: adaptive boosting of decision trees (AdaBoost), recurrent neural network ...
详细信息
A chatbot is an intelligent agent that developed based on Natural language processing (NLP) to interact with people in a natural language. The development of multiple deep NLP models has allowed for the creation ...
详细信息
Medical notes contain valuable information about patient conditions, treatments, and progress. Extracting symptoms from these unstructured notes is crucial for clinical research, population health analysis, and decisi...
Medical notes contain valuable information about patient conditions, treatments, and progress. Extracting symptoms from these unstructured notes is crucial for clinical research, population health analysis, and decision support systems. Traditional manual methods are time-consuming, but recent advances in natural language processing (NLP) and machine learning offer automated solutions. This article presents a novel approach that combines NLP techniques, such as conditional random fields (CRF) and transformer-based architectures. The proposed method demonstrates effective symptom extraction from medical notes, overcoming challenges such as varied terminologies and linguistic nuances. The study utilizes a dataset of Russian medical records, transforming it into a tabular format for training and employing unique tokenization algorithms for different models. Among the evaluated models, RuBERT achieved the highest accuracy of 91%, indicating its strong performance on the test dataset. SBERT exhibited the highest precision and F1 score, suggesting its effectiveness in accurately identifying specific sequence labels.
暂无评论