arXiv subjects

Ian Barnett

Publications and source records attributed to Ian Barnett.

22 records · Page 2Linked to original sources

arXiv2018

Inferring Mobility Measures from GPS Traces with Missing Data

With increasing availability of smartphones with GPS capabilities, large-scale studies relating individual-level mobility patterns to a wide variety of patient-centered outcomes, from mood disorders to surgical recovery, are becoming a reality. Similar past studies have been small in scale and have provided wearable GPS devices to subjects. These devices typically collect mobility traces continuously without significant gaps in the data, and consequently the problem of data missingness has been safely ignored. Leveraging subjects' own smartphones makes it possible to scale up and extend the duration of these types of studies, but at the same time introduces a substantial challenge: to preserve a smartphone's battery, GPS can be active only for a small portion of the time, frequently less than $10\%$, leading to a tremendous missing data problem. We introduce a principled statistical approach, based on weighted resampling of the observed data, to impute the missing mobility traces, which we then summarize using different mobility measures. We compare the strengths of our approach to linear interpolation, a popular approach for dealing with missing data, both analytically and through simulation of missingness for empirical data. We conclude that our imputation approach better mirrors human mobility both theoretically and over a sample of GPS mobility traces from 182 individuals in the Geolife data set, where, relative to linear interpolation, imputation resulted in a 10-fold reduction in the error averaged across all mobility features.

stat.ME↗

arXiv2016

Feature-Based Classification of Networks

Network representations of systems from various scientific and societal domains are neither completely random nor fully regular, but instead appear to contain recurring structural building blocks. These features tend to be shared by networks belonging to the same broad class, such as the class of social networks or the class of biological networks. At a finer scale of classification within each such class, networks describing more similar systems tend to have more similar features. This occurs presumably because networks representing similar purposes or constructions would be expected to be generated by a shared set of domain specific mechanisms, and it should therefore be possible to classify these networks into categories based on their features at various structural levels. Here we describe and demonstrate a new, hybrid approach that combines manual selection of features of potential interest with existing automated classification methods. In particular, selecting well-known and well-studied features that have been used throughout social network analysis and network science and then classifying with methods such as random forests that are of special utility in the presence of feature collinearity, we find that we achieve higher accuracy, in shorter computation time, with greater interpretability of the network classification results.

cs.SI↗

arXiv2016

Social and Spatial Clustering of People at Humanity's Largest Gathering

Macroscopic behavior of scientific and societal systems results from the aggregation of microscopic behaviors of their constituent elements, but connecting the macroscopic with the microscopic in human behavior has traditionally been difficult. Manifestations of homophily, the notion that individuals tend to interact with others who resemble them, have been observed in many small and intermediate size settings. However, whether this behavior translates to truly macroscopic levels, and what its consequences may be, remains unknown. Here, we use call detail records (CDRs) to examine the population dynamics and manifestations of social and spatial homophily at a macroscopic level among the residents of 23 states of India at the Kumbh Mela, a 3-month-long Hindu festival. We estimate that the festival was attended by 61 million people, making it the largest gathering in the history of humanity. While we find strong overall evidence for both types of homophily for residents of different states, participants from low-representation states show considerably stronger propensity for both social and spatial homophily than those from high-representation states. These manifestations of homophily are amplified on crowded days, such as the peak day of the festival, which we estimate was attended by 25 million people. Our findings confirm that homophily, which here likely arises from social influence, permeates all scales of human behavior.

stat.AP↗

arXiv2015

Change Point Detection in Correlation Networks

Many systems of interacting elements can be conceptualized as networks, where network nodes represent the elements and network ties represent interactions between the elements. In systems where the underlying network evolves in time, it is useful to determine the points in time where the network structure changes significantly as these may correspond also to functional change points. We propose a method for detecting these change points in correlation networks that, unlike previous change point detection methods designed for time series data, requires no distributional assumptions. We investigate the difficulty of change point detection near the boundaries of data in correlation networks and demonstrate the power of our method and a competing method through simulation. We also show the generalizable nature of our method by applying it to stock price data as well as fMRI data.

stat.ME↗