322,857 research outputs found
Epidemic spreading phenomena on a scale-free network with time-varying transmission rate due to social responses
No full textIn this paper, we study and simulate the effect of individual social responses, as a collective factor, on the epidemic spreading processes. We formally define the problem based on the traditional SI and SIR compartmental models considering the time-varying infection probability dependent on the social responses. In this study, models of generic and special case scenarios are developed. While in the generic case the effective parameter of behavioral response is demonstrated as one collective factor, in the special case the behavioral response is assumed as the combination of two collective factors: social cost and transfer rate of social awareness. With social cost, we refer to the costs incurred by a certain population to prevent or mitigate an epidemic. With transfer rate of social awareness, we describe the averaged rate of received information and knowledge regarding a disease that individuals hold and make use to avoid negative consequences. We show that, while in both SI and SIR models the density of infected agents grows exponentially during the initial time steps, the inclusion of our models of social responses, either generic or special one, leads to mitigation of the spreading. As a result of both generic and special cases, the density of infected agents in the stationary state and the maximum number of infected agents decrease according to power-law functions for different values of collective factors. In the special case results, we also witnessed significant changes in the slope of decreasing trends of stationary density of states happening for a critical value of transfer rate of social awareness, approximately at about the inverse of the time interval of transmission rate update. With this result, we point out that increasing the transfer rate of social awareness to about this critical point outperforms any slight increase in social cost in reducing the number of infected agents
An NLP-based statistical reporting methodology applied to court decisions
No full textNatural Language Processing (NLP) algorithms
have significantly advanced the capabilities of understanding,
processing and generating human language. However, one persis-
tent challenge in NLP is the problem of uncertainty, due e.g. to the
inherent complexity of human language, variations in language
usage across different contexts and domains, and the presence of
noisy or incomplete data. In this work we take in consideration
this problem for statistics derived from court documents by NLP
systems
The Unknown of the Pandemic: An Agent-Based Model of Final Phase Risks
No full textLifting social restrictions is one of the most critical decisions that public health authorities have to face during a pandemic such as COVID-19. This work focuses on the risk associated with such a decision. We have called the period from the re-opening decision to epidemic expiration the ’final epidemic phase’, and considered the critical epidemic conditions which could possibly emerge in this phase. The factors we have considered include: the proportion of asymptomatic cases, a mitigation strategy based on testing and the average duration of infectious states. By assuming hypothetical configurations at the time of the re-opening decision and the partial knowledge concerning epidemic dynamics available to public health authorities, we have analyzed the risk of the re-opening decision based on possibly unreliable estimates. We have presented a discrete-time stochastic model with state-dependent transmission probabilities and multi-agent simulations. Our results show the different outcomes produced by different proportions of undetected asymptomatic cases, different probabilities of asymptomatic cases detected and contained, and a multivariate analysis of risk based on the average duration of asymptomatic and contained states. Finally, our analysis highlights that enduring uncertainty, typical of this pandemic, requires a risk analysis approach to complement epidemiological studies
Three-particle Bell-like inequalities under Lorentz transformations
No full textWe study the effects of Lorentz transformations on three-particle nonlocal system states (GHZ and W) of spin 1/2 particles, using the Pauli spin operator and a three-particle generalization of Bell's inequality, introduced by Svetlichny. In our setup, the moving and laboratory frames used the (same) set of measurement directions that maximally violate Svetlichny's inequality in the laboratory frame. We also investigate the behavior of Mermin's and Collins' inequalities. We find that, regardless of the particles' type of entanglement, violation of Svetlichny's inequality in the moving frame is decreased by increasing the boost velocity and the energy of particles in the laboratory frame. In the relativistic regime, Svetlichny's inequality is a good criterion to investigate the nonlocality of the GHZ state. We also find that Mermin's and Collins' inequalities lead to reasonable predictions, in agreement with the behavior of the spin state, about nonlocality of the W state in the relativistic regime. Then, comparing our results with those in which Czachor's relativistic spin is used instead of the Pauli operator, we find that the results obtained by considering the Pauli spin operator are in better agreement with the behavior of spin state of the system in the relativistic information theory
Digital Health Data, a Way to Take Under Control the Quality During the Elaboration Processes
No full textIn this work we explore the uncertainty in measured attributes of events, correlating with other events happening in a near time window. The conceptual framework of the method is the event graph concept. Our objective is to filter out the unreliable events and obtain an optimized fraction of data, in order to compute decent statistics, to be able to evaluate the data and build derived measures. Developing our framework, we proceed with some experiments using data from wearable devices, both synthetic data sets and a public available benchmark
Correlation and pattern detection in event networks
No full textEvents happening at defined moments in time and involving specific entities from a social or physical system can be organized in networks or graphs. The study of such event graphs may reveal causal relations between subsequent events or compound events that we define as “typed events”. Moreover, characteristic sequences of events or patterns can arise in consequence of phenomena affecting the system. Methods to build the event graph and to search for the typed events and their significance are described in detail. An embedding strategy to encode typed events in low dimensional vectors is defined, and both supervised and unsupervised learning is applied to search for meaningful patterns. Experiments have been conducted using data from a real investigation and some synthetic data
Technologies and Strategies for Continuous Learning through Electronic Health Records Data
No full textAchieving a comprehensive view of a patient’s health using data from
Electronic Health Record systems requires the use of advanced analytics. However,
effectively managing and curating this data requires carefully designed workflows.
While digitization and standardization enable continuous health monitoring, issues
such as missing data values and technical glitches can jeopardize data consistency
and timeliness. On the other hand, the Efficiency in processing the large volume
of data from disparate sources generated by the healthcare industry is critical. In
this chapter, we try to provide an overview of how distributed computing and
Artificial Intelligence can be used in the context of smart healthcare and big data
in practical use cases, enabling insights to improve patient care. In addition, we
propose a workflow for developing prognostic models that uses the SMART BEAR
infrastructure and leverages the capabilities of the Big Data Analytics engine to
standardize and harmonize data. Our workflow improves data quality by evaluating
different imputation algorithms and selecting the one that preserves the distribution
and correlation of features similar to the original data. We applied this workflow to a
subset of data in the SMART BEAR repository and evaluated its impact on predicting
future health conditions, such as cardiovascular disease and mild depression. We
also explored the potential for model validation by clinicians in the SMART BEAR
project, the transfer of subsequent actions within the decision support system, and
the estimation of the required number of data points
Graph embeddings in criminal investigation: towards combining precision, generalization and transparency
No full textGraph Embeddings in Criminal Investigationn: Extending the Scope of Enquiry Protocols
No full textKnowledge graphs are exploited in criminal investigation to integrate heterogeneous data sources and scale up the operational efficiency of enquiry protocols. Using a declarative perspective, protocols can be viewed as a set of data ingestion procedures and nested exact queries. This meets the probating nature of procedural justice that has to proceed from established facts. At the same time, the exact specification of queries represents a limit for enquiry protocols that can exclusively retrieve those facts in adherence to the designed queries. We then investigated the use of graph em-beddings procedures to extend the scope of a protocol by returning sub-graphs partially matching to its specification. Because exploring the entire set of sub-graphs quickly become computationally intractable, we developed an approach based on a hierarchical filtering procedure. A controlled experiment we executed has shown the feasibility of our approach
Data management for continuous learning in EHR systems
No full textTo gain a comprehensive understanding of a patient’s health, advanced analytics must be applied to the data collected by electronic health record (EHR) systems. However, managing and curating this data requires carefully designed workflows. While digitalization and standardization enable continuous health monitoring, missing data values and technical issues can compromise the consistency and timeliness of the data. In this paper, we propose a workflow for developing prognostic models that leverages the SMART BEAR infrastructure and the capabilities of the Big Data Analytics (BDA) engine to homogenize and harmonize data points. Our workflow improves the quality of the data by evaluating different imputation algorithms and selecting one that maintains the distribution and correlation of features similar to the raw data. We applied this workflow to a subset of the data stored in the SMART BEAR repository and examined its impact on the prediction of emerging health states such as cardiovascular disease and mild depression. We also discussed the possibility of model validation by clinicians in the SMART BEAR project, the transmission of subsequent actions in the decision support system, and the estimation of the required number of data points
- …
