Joseph D. Skufca

Machine learning, medical diagnosis, and biomedical engineering research - commentary

A large number of papers are appearing in the biomedical engineering literature that describe the use of machine learning techniques to develop classifiers for detection or diagnosis of disease. However, the usefulness of this approach in developing clinically validated diagnostic techniques so far has been limited and the methods are prone to overfitting and other problems which may not be immediately apparent to the investigators. This commentary is intended to help sensitize investigators as well as readers and reviewers of papers to some potential pitfalls in the development of classifiers, and suggests steps that researchers can take to help avoid these problems. Building classifiers should be viewed not simply as an add-on statistical analysis, but as part and parcel of the experimental process. Validation of classifiers for diagnostic applications should be considered as part of a much larger process of establishing the clinical validity of the diagnostic technique.

The effects of pick density on order picking areas with narrow aisles

The cost and service performance of an order fulfillment center are determined partly by how workers are organized into an order picking system. One common approach is batch picking, in which workers circumnavigate a picking area with other workers, gathering items on a pick list. In some systems with high space utilization, narrow aisles prohibit workers from passing one another when in the same aisle, and this leads to congestion. We build analytical and simulation models of these systems to investigate their behavior under different levels of activity. Among other things, our results suggest that when the system is busier and pick density is high (that is, when workers stop often to make picks) congestion is less of a problem and workers are more productive.

Amplitude Death Solutions for Stabilization of DC Microgrids With Instantaneous Constant-Power Loads

Constant-power loads on dc microgrids create a destabilizing effect on the circuit that can lead to severe voltage and frequency oscillations. Amplitude death is a coupling induced stabilization of the fixed point of a dynamical system. This paper applies amplitude death methods to the stabilization problem in this constant-power setting. The amplitude death methods provide an open-loop control solution to stabilize the system. Two methods, one using delay, the other using circuit heterogeneity, are examined. Each method is demonstrated through numerical simulations.

Portraits of complex networks

We propose a method for characterizing large complex networks by introducing a new matrix structure, unique for a given network, which encodes structural information; provides useful visualization, even for very large networks; and allows for rigorous statistical comparison between networks. Dynamic processes such as percolation can be visualized using animation.

Control entropy identifies differential changes in complexity of walking and running gait patterns with increasing speed in highly trained runners

Regularity statistics have been previously applied to walking gait measures in the hope of gaining insight into the complexity of gait under different conditions and in different populations. Traditional regularity statistics are subject to the requirement of stationarity, a limitation for examining changes in complexity under dynamic conditions such as exhaustive exercise. Using a novel measure, control entropy (CE), applied to triaxial continuous accelerometry, we report changes in complexity of walking and running during increasing speeds up to exhaustion in highly trained runners. We further apply Karhunen-Loeve analysis in a new and novel way to the patterns of CE responses in each of the three axes to identify dominant modes of CE responses in the vertical, mediolateral, and anterior/posterior planes. The differential CE responses observed between the different axes in this select population provide insight into the constraints of walking and running in those who may have optimized locomotion. Future comparisons between athletes, healthy untrained, and clinical populations using this approach may help elucidate differences between optimized and diseased locomotor control.

k Workers in a Circular Warehouse: A Random Walk on a Circle, without Passing

We consider the problem of stochastic flow of multiple particles traveling on a closed loop, with a constraint that particles move without passing. We use a Markov chain description that reduces the problem to a generalized random walk on a hyperplane (with boundaries). By expressing positions via a moving reference frame, the geometry of the no-passing criteria is greatly simplified, with the resultant condition expressible as the coordinate system planes which bound the first orthant. To determine state transition probabilities, we decompose transitions into independent events and construct a digraph representation in which calculating transition probability is reduced to a shortest path determination on the digraph. The resultant decomposition digraph is self-converse, and we exploit that property to establish the necessary symmetries to find the stationary density for the process.

Analysis Still Matters: A Surprising Instance of Failure of Runge-Kutta-Felberg ODE Solvers

This paper provides a nice example to illustrate that without supporting analysis, a numerical simulation may lead to incorrect conclusions. We explore a pedagogical example of failure of Runge--Kutta--Felberg (RKF) algorithms for a simple dynamical system that models the coupling of two oscillators. Although the system appears to be well-behaved, the explicit RKF solvers provide erratic numerical solutions. The mode of failure is based in a period-doubling route to chaos due to the existence of stable linear solutions in the problem.

A statistical approach to the use of control entropy identifies differences in constraints of gait in highly trained versus untrained runners.

Control entropy (CE) is a complexity analysis suitable for dynamic, non-stationary conditions which allows the inference of the control effort of a dynamical system generating the signal. These characteristics make CE a highly relevant time varying quantity relevant to the dynamic physiological responses associated with running. Using High Resolution Accelerometry (HRA) signals we evaluate here constraints of running gait, from two different groups of runners, highly trained collegiate and untrained runners. To this end,we further develop the control entropy (CE) statistic to allow for group analysis to examine the non-linear characteristics of movement patterns in highly trained runners with those of untrained runners, to gain insight regarding gaits that are optimal for running. Specifically, CE develops response time series of individuals descriptive of the control effort; a group analysis of these shapes developed here uses Karhunen Loeve Analysis (KL) modes of these time series which are compared between groups by application of a Hotelling T² test to these group response shapes. We find that differences in the shape of the CE response exist within groups, between axes for untrained runners (vertical vs anterior-posterior and mediolateral vs anterior-posterior) and trained runners (mediolateral vs anterior-posterior). Also shape differences exist between groups by axes (vertical vs mediolateral). Further, the CE, as a whole, was higher in each axis in trained vs untrained runners. These results indicate that the approach can provide unique insight regarding the differing constraints on running gait in highly trained and untrained runners when running under dynamic conditions. Further, the final point indicates trained runners are less constrained than untrained runners across all running speeds.

Are College Students’ Assessments of Threat Shaped by the Dangers of Their Childhood Environment?

Humans internalize environmental cues of mortality risk at an early age, which influences subsequent risk perceptions and behavior. In this respect, an individual’s current risk assessment may be viewed as an adaptive response to the dangers present within his or her early local environment. Here we examine the relationship between several variables indicating threat within an individual’s early environment (e.g., prevalence of violent and property crimes, registered sex offenders) and their perception of crime risk within both the childhood and current adult environments. We recruited a group of 657 students who hail from diverse geographic backgrounds to provide the zip code location of their childhood residence along with subjective ratings of danger of that and their current location, which enabled us to compare their ratings of risk/danger with the federally reported crime statistics of each setting. Our results indicate that the early prevalence of registered sex offenders indeed influences an individual’s risk perception in adulthood, and that these factors have a differential effect on males and females. Our findings provide support for the theory that early environmental factors signaling danger affect how individuals assess risk within their adult environment.

The Problem of False Discovery: Many Scientific Results Can't Be Replicated, Leading to Serious Questions about What's True and False in the World of Research

Is there a Cheshire Cat in science? One might believe so, given the many published scientific discoveries that cannot be independently reproduced. The ?replication crisis? in science has become a widely discussed issue among scientists and the lay media and even has its own entry in Wikipedia.