Page 37 - Invited Paper Session (IPS) - Volume 1
P. 37
IPS35 Dinov I.D. et al.
Figure 4: DataSifter obfuscation – trade-offs between privacy protection and
preservation of data utility.
In addition, we use established model-based and model-free techniques
to interrogate the data (Dinov 2016, Dinov 2016, Dinov 2018, Gao, Sun et al.
2018, Kalinin, Allyn-Feuer et al. 2018, Marino, Xu et al. 2018, Tang, Gao et al.
2018, Zhao, Matloff et al. 2018). These include both confirmatory (hypothesis
driven) and exploratory (visual analytics) inferential techniques to extract
knowledge, identify patterns, forecast trends, and forecast univariate
outcomes of interest and derived computed phenotypes.
3. Results
Open-science relies heavily on data sharing, findability, accessibility,
interoperability, and reusability (FAIR) (Wilkinson, Dumontier et al. 2016),
open-source development (Feller and Fitzgerald 2002), and transdisciplinary
cooperation (Kreps and Maibach 2008, Dinov 2018). Figure 5 presents some
examples of recent results illustrating the power of advanced mathematical
modelling techniques, statistical inferential methods, and machine learning
strategies to analyse complex, multisource, heterogeneous, and incomplete
datasets.
26 | I S I W S C 2 0 1 9
