Computing and Informatics (E-Journal - Institute of Informatics, SAS, Bratislava)
Not a member yet
1506 research outputs found
Sort by
Incremental Learning Method for Data with Delayed Labels
Get PDFMost research on machine learning tasks relies on the availability of true labels immediately after making a prediction. However, in many cases, the ground truth labels become available with a non-negligible delay. In general, delayed labels create two problems. First, labelled data is insufficient because the label for each data chunk will be obtained multiple times. Second, there remains a problem of concept drift due to the long period of data. In this work, we propose a novel incremental ensemble learning when delayed labels occur. First, we build a sliding time window to preserve the historical data. Then we train an adaptive classifier by labelled data in the sliding time window. It is worth noting that we improve the TrAdaBoost to expand the data of the latest moment when building an adaptive classifier. It can correctly distinguish the wrong types of source domain sample classification. Finally, we integrate the various classifiers to make predictions. We apply our algorithms to synthetic and real credit scoring datasets. The experiment results indicate our algorithms have superiority in delayed labelling setting
A Connected Mobility Scheme for Taxi Supply-Demand Balancing in a Smart City Context
Get PDFIn this paper we present the preliminary results of simulation-based experiments of an integrated scheme that has been proposed to control taxi supply-demand imbalance in the context of a smart city with multiple taxi operators and using Connected Mobility. We particularly explore the difference between centralized and decentralized implementations of the scheme as well as between collaborative and competitive attitudes of connected taxis. Our results show that by sharing knowledge about supply-demand imbalance and adopting a collaborative attitude, connected taxi systems can improve the performance of the service across a city by achieving a better supply-demand service balancing while improving their profits
Modular Design and Adaptive Control of Urban Signalized Intersections Systems Using Synchronized Timed Petri Nets
Get PDFTraffic flow at urban intersections varies randomly during the day. It depends on several dynamic factors and requires efficient regulation and flexible control strategies in particular for traffic light regulation. The proposed strategy allows managing the green light time autonomously. The dynamic behavior of traffic signals at intersections can be seen as a discrete event system. Through this paper, a modular Timed Synchronized Petri Net (TSPN) model is developed and a real-time adaptive control strategy of urban signalized intersections is proposed. The control is shared between two communicant actors. The master-slaves approach is adopted in this control strategy. The master (controller) decides the next phase to be served with green light and its duration. While, the slaves (TSPN modules) control the traffic signals displays, phases transitions, and model traffic flow fluctuations. Thanks to the used modularity approach, the developed models reduce the system complexity in terms of combinatorial explosion, and they could be adapted easily for any real intersection. Using the developed models, some interesting properties of the system are checked, and some simulations are performed and analyzed in order to validate the proposed control approach
Application of the Fuzzy Model Theory for Modeling QA-Systems
Get PDFThe work is devoted to the description of the question-answer system QA-RiskPanel, which provides means of determining and forecasting the risks related to computer attacks. The QA-RiskPanel system uses a constantly updated database of previous computer attacks as a source of knowledge. We thus guarantee the most up-to-date risk prediction. The ontological approach to the formalization of the object domain allows the analysis of risks at various levels of specification/generalization. In this paper we provide a model-theoretic formalization of the Knowledge Base of the described object domain. Then we describe the classification of question types, which are probabilistic in this system. Finally we present algorithms for finding the answers to all question types of our classification
Enhanced Critical Node Detection in Social Networks
Get PDFIn this paper, we investigate the popular centrality-based approaches to find a set of critical nodes whose deletion causes the most disconnectivity in the network. Demonstrating the weak points of these approaches which only consider a ranking factor, we propose an Enhanced Critical Node Detection (ECND) method which can work with any kind of ranking score by considering the structure of a network. We have designed a set of experiments using 24 different artificial and real-world networks, varying in the number of vertices and number of edges. Using two different objective functions including the number of connected components and the weighted average size of the connected components, experimental results show outperformance of ECND in comparison to all 8 other methods
Context-Aware Music Recommendation with Metadata Awareness and Recurrent Neural Networks
Get PDFDay by day, music streaming services grow the volume of data on the internet. To help the users to find songs that fit their interests, music recommender systems can be used to filter a large number of songs according to the preference of the user. However, the context in which the users listen to songs must be taken into account, which justifies the usage of context-aware recommender systems. Although there are some works about context-aware music recommender systems, there is a lack of automatic techniques for extracting contextual information for these systems. Thus, the goal of this work is to propose two methods to acquire contextual information (represented by embeddings) for each song, given the sequence of songs that each user has listened to. The first method, called Metadata-Aware, uses tags and genres to enrich the embeddings with additional information. The second method, called Dual Recurrent Neural Network, uses such a network to improve the embeddings generated from long sequences of songs. The embeddings generated by both methods were evaluated with four context-aware music recommender systems in two datasets. The results showed that the embeddings, obtained by our proposals, present better results than the state-of-the-art method proposed in the literature (in some cases with gains of more than 100 %). Finally, the experiments also showed that our second method provides better results than the first one
Artificial Intelligence (AI) Model: Adaptive Neuro-Fuzzy Inference System (ANFIS) for Diagnosis of COVID-19 Influenza
Get PDFThe COVID-19 influenza became a curse on the world. It has been around for two years, so no one needs to make a big introduction of it. It has became a significant challenge around the world. Owing to this, we made dynamic networks using an amalgamating of fuzzy logic and neural networks for the prediction of sufferers of COVID-19. These hybrid networks serve for the assessment of the COVID-19 victims and usefully serve for the assessment of the medical resources needed for future victims. This manuscript proposed Sugeno Adaptive Neuro-Fuzzy Inference System (SANFIS) prediction model for COVID-19 prediction in Andhra Pradesh, India. We gathered data on positive COVID-19 sufferers in Andhra Pradesh for this purpose. The data can be separated into three categories: training set, testing set and checking set. We have utilized Root Mean Square Deviation (RMSD) for prediction precision. If the prediction model has a lower RMSD value, it is regarded as the best forecast. In this study, we concluded that the 3 Triangular MFns for each input were excellent with the extreme precision for all of the districts based on our expertise. In the end, we deployed seven SANFIS replicas in Andhra Pradesh, but we discovered that SANFIS6 and SANFIS7 provided excellent COVID-19 prediction results. These findings will assist the government, healthcare agencies, and medical organizations in planning for future COVID-19 victims' medical requirements. These sorts of Sugeno Adaptive Neuro-Fuzzy Inference System (SANFIS) prediction models based on Artificial Intelligence (AI) will be beneficial in overcoming the COVID-19
Review of Smart Contracts for Cloud-Based Manufacturing
Get PDFCloud-based manufacturing is taking shape, and many industries seem interested to make the transition to it. Developing blockchain solutions for trusted computing is also taking its roots. Developing a blockchain-based solution for cloud-based manufacturing systems is a field that is new but also faces limitations and a lack of case studies. Smart contracts are one part of the solution which deals with making blockchain successful in cloud-based manufacturing. As we move towards smart contracts design and development for cloud-based manufacturing, there is no complete survey of smart contract and cloud manufacturing that can highlight critical, challenging issues and limitations. Most of the work found in smart contracts is mostly financial and notary-centric applications. On the cloud manufacturing side, most of the literature deals with Internet of Things (IoT) and cloud computing systems. Therefore, there is a need to study the best practices to start manufacturing supported by blockchain smart contracts. We conducted a scoping review for smart contracts for cloud manufacturing to address the problem mentioned above. We studied the latest case studies and concepts in data extracted from digital libraries and online repositories. Furthermore, we follow the relevance and acceptance criteria of research articles for inclusion and exclusion from this work. This paper focuses on blockchain systems, smart contracts and architecture, smart contracts in the cloud, and the IoT environment. Furthermore, we tried to bridge design and implementation details for readers to understand the patterns that can replicate for cloud-based manufacturing systems
Semantic Segmentation of Text Using Deep Learning
Get PDFGiven a text, can we segment it into semantically coherent sections in an automatic way? Can we detect the semantic boundaries, if we know how many they are? Can we determine how many semantically distinct sections are in the text? These are the questions we address in this paper. To respond, we use the Bidirectional Encoder Representation from Transformer (BERT) to analyze the text and evaluate a function that we call local incoherence, which we expect to show maxima at the points where a semantic boundary is detected. Our results, although preliminary, are encouraging and suggest that our approach can be successfully applied. However, they are quite sensitive with respect to the text quality, as it happens in the case in which the text is derived from an audio stream via Automatic Speech Recognition techniques
Reducing the Effect of Imbalance in Text Classification Using SVD and GloVe with Ensemble and Deep Learning
Get PDFDue to the recent escalation in the amount of text data available and used online, text classification has become a staple for data analysts when extracting relevant information. Yet, machine learning algorithms are susceptible to biases when implemented on any large-scale automated task, especially in text analysis. With the popularization of newer branches of study emerging from the field of machine learning – such as ensemble and deep learning – we must analyze the potential pitfalls in the common experimental setup centered around learning algorithms. Imbalance in text data is one such pitfall – when data is not equally distributed across all categories in a dataset, it can influence and undermine the classification of underrepresented categories. In our research, we have proposed several techniques and unique approaches to tackle this obstacle. We prepared four datasets of varying degrees of imbalance to conduct our experimentation. We proved that feature extraction techniques singular value decomposition (SVD) and GloVe are the key to reducing the effect of imbalance in text classification, especially in ensemble and deep learning. Using the result of our research, we have also proposed a modified ensemble classifier that can classify imbalanced and balanced data alike