Insight Research Collection

Permanent URI for this collection

At Insight we undertake high impact research in data analytics that has significant impact on industry and society by enabling better decision making.

Insight brings together leading Irish academics from 5 of Ireland's leading research centres (DERI, CLARITY, CLIQUE, 4C, TRIL), previously established by Science Foundation Ireland (SFI) and the Irish Industrial Development Authority (IDA), in key areas of priority research including:

  • The Semantic Web
  • Sensors and the Sensor Web
  • Social network analysis
  • Decision Support and Optimization
  • Connected Health

For more information, please visit the official website.

Browse

Recent Submissions

Now showing 1 - 5 of 539
  • Publication
    Combining biomarker and food intake data
    Recent developments in biomarker discovery have demonstrated that combining biomarkers with self-reported intake data has the potential to improve estimation of food intake. Here, statistical methods for combining biomarker and self-reported food intake data are discussed. The calibration equations method is a widely applied method that corrects for measurement error in self-reported food intake data through the use of biomarker data. The method is outlined and illustrated through an example where citrus intake is estimated. In order to estimate stable calibration equations, a simulation-based framework is delineated which estimates the percentage of study subjects from whom biomarker data is required. The method of triads is frequently used to assess the validity of self-reported food intake data by combining it with biomarker data. The method is outlined and sensitivity to its underlying assumptions is illustrated through simulation studies.
      20
  • Publication
    Inferring food intake from multiple biomarkers using a latent variable model
    (Institute of Mathematical Statistics, 2021-12) ; ;
    Metabolomic based approaches have gained much attention in recent years due to their promising potential to deliver objective tools for assessment of food intake. In particular, multiple biomarkers have emerged for single foods. However, there is a lack of statistical tools available for combining multiple biomarkers to quantitatively infer food intake. Furthermore, there is a paucity of approaches for estimating the uncertainty around biomarker-based inferred intake. Here, to estimate the relationship between multiple metabolomic biomarkers and food intake in an intervention study conducted under the A-DIET research programme, a latent variable model, multiMarker, is proposed. The multiMarker model integrates factor analytic and mixture of experts models: the observed biomarker values are related to intake which is described as a continuous latent variable which follows a flexible mixture of experts model with Gaussian components. The multiMarker model also facilitates inference on the latent intake when only biomarker data are subsequently observed. A Bayesian hierarchical modelling framework provides flexibility to adapt to different biomarker distributions and facilitates inference of the latent intake along with its associated uncertainty. Simulation studies are conducted to assess the performance of the multiMarker model, prior to its application to the motivating application of quantifying apple intake.
      15Scopus© Citations 1
  • Publication
    Regional Knowledge Spaces: The Interplay of Entry-Relatedness and Entry-Potential for Technological Change and Growth
    (University College Dublin. Spatial Dynamics Lab, 2021-09-20) ; ; ;
    This paper aims to uncover the mechanism of how the network properties of regional knowledge spaces contribute to technological change from the perspective of regional knowledge entry-relatedness and regional knowledge entry-potential. Entry-relatedness, which has been previously employed to investigate the technology evolution of regional economies, is advanced by introducing a knowledge gravity model. The entry-potential of a newly acquired regional specialisation has been largely ignored in the relevant literature; surprisingly given the high relevance that is attributed to the recombination potential of new capabilities. In other words, just adding new knowledge domains to a system is not sufficient alone, it really depends on how these fit into the existing system and thus can generate wider economic benefits. Based on an empirical analysis of EU Metro and non-Metro regions from 1981 to 2015, we find that entry-relatedness has a significant negative association with novel inventive activities, while entry-potential has a significant positive association with the development of novel products and processes of economic value. This highlights that regions’ capacity to venture into high-potential areas of technological specialization in the knowledge space outperforms purely relatedness driven diversification that is frequently promoted in the relevant literature.
      143
  • Publication
    Bayesian Inference, Model Selection and Likelihood Estimation using Fast Rejection Sampling: The Conway-Maxwell-Poisson Distribution
    (International Society for Bayesian Analysis, 2021-09) ;
    Bayesian inference for models with intractable likelihood functions represents a challenging suite of problems in modern statistics. In this work we analyse the Conway-Maxwell-Poisson (COM-Poisson) distribution, a two parameter generalisation of the Poisson distribution. COM-Poisson regression modelling allows the flexibility to model dispersed count data as part of a generalised linear model (GLM) with a COM-Poisson response, where exogenous covariates control the mean and dispersion level of the response. The major difficulty with COM-Poisson regression is that the likelihood function contains multiple intractable normalising constants and is not amenable to standard inference and Markov Chain Monte Carlo (MCMC) techniques. Recent work by Chanialidis et al. (2018) has seen the development of a sampler to draw random variates from the COM-Poisson likelihood using a rejection sampling algorithm. We provide a new rejection sampler for the COM-Poisson distribution which significantly reduces the central processing unit (CPU) time required to perform inference for COM-Poisson regression models. An extension of this work shows that for any intractable likelihood function with an associated rejection sampler it is possible to construct unbiased estimators of the intractable likelihood which proves useful for model selection or for use within pseudo-marginal MCMC algorithms (Andrieu and Roberts, 2009). We demonstrate all of these methods on a real-world dataset of takeover bids.
      72Scopus© Citations 3
  • Publication
    Overlap Training to Mitigate Inconsistencies Caused by Image Tiling in CNNs
    (Springer, 2020-12-08) ; ; ;
    This paper focuses on the problem of inconsistent predictions of modern convolutional neural networks (CNN) at patch (i.e. sub-image) boundaries. Limited by the graphics processing unit (GPU) resources, image tiling and stitching countermeasure have been applied for most megapixel images, that is, cutting images into overlapping tiles as CNN input, and then stitching CNN outputs together. However, we found that stitched (i.e. recovered) predictions have discontinuous grid-like noise. We propose a simple yet efficient overlap training framework to mitigate the inconsistent prediction at patch boundaries without changing the model architecture while improving the stability, robustness of the model. We have applied our solution to various CNNs (such as U-Net, DeepLab, RCF) and tested them on two real-world datasets. Extensive experiments suggest that the new framework is sufficient in reducing inconsistency and outperform these countermeasures. The source code and coloured figures are made publicly available online at: https://github.com/anyuzoey/Overlap-Training.git.
      196Scopus© Citations 2