Currently, autonomous vehicles are being evaluated by how well they interact with humans without evaluating how well humans interact with them. Since people are not going to unanimously switch over to using autonomous vehicles, attention must be given to how well these new vehicles signal intent to human drivers from the driver’s point of view. Ineffective communication will lead to unnecessary discomfort among drivers caused by an underlying uncertainty about what an autonomous vehicle is or isn’t about to do. Recent studies suggest that humans tend to fixate on areas of higher uncertainty so scenarios that have a higher number of vehicle fixations can be reasoned to be more uncertain. We provide a framework for measuring human uncertainty and use the framework to measure the effect of empathetic vs non-empathetic agents. We used a simulated driving environment to create recorded scenarios and manipulate the autonomous vehicle to include either an empathetic or non-empathetic agent. The driving interaction is composed of two vehicles approaching an uncontrolled intersection. These scenarios were played to twelve participants while their gaze was recorded to track what the participants were fixating on. The overall intent was to provide an analytical framework as a tool for evaluating autonomous driving features; and in this case, we choose to evaluate how effective it was for vehicles to have empathetic behaviors included in the autonomous vehicle decision making. A t-test analysis of the gaze indicated that empathy did not in fact reduce uncertainty although additional testing of this hypothesis will be needed due to the small sample size.
In the weeks following the first imported case of Ebola in the U. S. on September 29, 2014, coverage of the very limited outbreak dominated the news media, in a manner quite disproportionate to the actual threat to national public health; by the end of October, 2014, there were only four laboratory confirmed cases of Ebola in the entire nation. Public interest in these events was high, as reflected in the millions of Ebola-related Internet searches and tweets performed in the month following the first confirmed case. Use of trending Internet searches and tweets has been proposed in the past for real-time prediction of outbreaks (a field referred to as “digital epidemiology”), but accounting for the biases of public panic has been problematic. In the case of the limited U. S. Ebola outbreak, we know that the Ebola-related searches and tweets originating the U. S. during the outbreak were due only to public interest or panic, providing an unprecedented means to determine how these dynamics affect such data, and how news media may be driving these trends.
Methodology
We examine daily Ebola-related Internet search and Twitter data in the U. S. during the six week period ending Oct 31, 2014. TV news coverage data were obtained from the daily number of Ebola-related news videos appearing on two major news networks. We fit the parameters of a mathematical contagion model to the data to determine if the news coverage was a significant factor in the temporal patterns in Ebola-related Internet and Twitter data.
Conclusions
We find significant evidence of contagion, with each Ebola-related news video inspiring tens of thousands of Ebola-related tweets and Internet searches. Between 65% to 76% of the variance in all samples is described by the news media contagion model.
Seroepidemiological studies before and after the epidemic wave of H1N1-2009 are useful for estimating population attack rates with a potential to validate early estimates of the reproduction number, R, in modeling studies.
Methodology/Principal Findings
Since the final epidemic size, the proportion of individuals in a population who become infected during an epidemic, is not the result of a binomial sampling process because infection events are not independent of each other, we propose the use of an asymptotic distribution of the final size to compute approximate 95% confidence intervals of the observed final size. This allows the comparison of the observed final sizes against predictions based on the modeling study (R = 1.15, 1.40 and 1.90), which also yields simple formulae for determining sample sizes for future seroepidemiological studies. We examine a total of eleven published seroepidemiological studies of H1N1-2009 that took place after observing the peak incidence in a number of countries. Observed seropositive proportions in six studies appear to be smaller than that predicted from R = 1.40; four of the six studies sampled serum less than one month after the reported peak incidence. The comparison of the observed final sizes against R = 1.15 and 1.90 reveals that all eleven studies appear not to be significantly deviating from the prediction with R = 1.15, but final sizes in nine studies indicate overestimation if the value R = 1.90 is used.
Conclusions
Sample sizes of published seroepidemiological studies were too small to assess the validity of model predictions except when R = 1.90 was used. We recommend the use of the proposed approach in determining the sample size of post-epidemic seroepidemiological studies, calculating the 95% confidence interval of observed final size, and conducting relevant hypothesis testing instead of the use of methods that rely on a binomial proportion.
Several past studies have found that media reports of suicides and homicides appear to subsequently increase the incidence of similar events in the community, apparently due to the coverage planting the seeds of ideation in at-risk individuals to commit similar acts.
Methods
Here we explore whether or not contagion is evident in more high-profile incidents, such as school shootings and mass killings (incidents with four or more people killed). We fit a contagion model to recent data sets related to such incidents in the US, with terms that take into account the fact that a school shooting or mass murder may temporarily increase the probability of a similar event in the immediate future, by assuming an exponential decay in contagiousness after an event.
Conclusions
We find significant evidence that mass killings involving firearms are incented by similar events in the immediate past. On average, this temporary increase in probability lasts 13 days, and each incident incites at least 0.30 new incidents (p = 0.0015). We also find significant evidence of contagion in school shootings, for which an incident is contagious for an average of 13 days, and incites an average of at least 0.22 new incidents (p = 0.0001). All p-values are assessed based on a likelihood ratio test comparing the likelihood of a contagion model to that of a null model with no contagion. On average, mass killings involving firearms occur approximately every two weeks in the US, while school shootings occur on average monthly. We find that state prevalence of firearm ownership is significantly associated with the state incidence of mass killings with firearms, school shootings, and mass shootings.
Throughout this project, I decided on a number of learning goals to consider it a success. I needed to learn how to use the supporting libraries that would help me to design this system. I also learned how to use the Twitter API, as well as create the infrastructure behind it that would allow me to collect large amounts of data for machine learning. I needed to become familiar with common machine learning libraries in Python in order to create the necessary algorithms and pipelines to make predictions based on Twitter data.
This paper details the steps and decisions needed to determine how to collect this data and apply it to machine learning algorithms. I determined how to create labelled data using pre-existing Botometer ratings, and the levels of confidence I needed to label data for training. I use the scikit-learn library to create these algorithms to best detect these bots. I used a number of pre-processing routines to refine the classifiers’ precision, including natural language processing and data analysis techniques. I eventually move to remotely-hosted versions of the system on Amazon web instances to collect larger amounts of data and train more advanced classifiers. This leads to the details of my final implementation of a user-facing server, hosted on AWS and interfacing over Gmail’s IMAP server.
The current and future development of this system is laid out. This includes more advanced classifiers, better data analysis, conversions to third party Twitter data collection systems, and user features. I detail what it is I have learned from this exercise, and what it is I hope to continue working on.