Browsing by Subject "regression"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Accessibility and Centrality Based Estimation of Pedestrian Activity and Safety in Urban Areas
    (2015-10) Murphy, Brendan
    The following thesis investigates the feasibility of using metrics of accessibility to jobs, betweenness centrality, and automobile traffic levels to estimate pedestrian behav- ior levels and automobile-pedestrian collision risks within an urban area. Multimodal count and crash report data from Minneapolis, Minnesota are used as a test of this scal- able, translatable modeling framework; multiple stepwise linear regression is performed to compile a set of explanatory variables from which to construct a predictive model of pedestrian movement. The existence of the Safety In Numbers (SIN) phenomenon is in- vestigated within both the raw and estimated pedestrian movement data; the SIN effect is the phenomenon where pedestrians are found to be safer from collisions, on average, when there are more pedestrians present in a given intersection, street, or area - that is, that the per-pedestrian risk of injury inflicted by drivers of automobiles decreases as a function of the increasing volume of pedestrian traffic. Economic accessibility, between- ness centrality, and Average Annual Daily Traffic (AADT) were found to be significant predictors of pedestrian traffic at intersections in Minneapolis, and the SIN effect was observed in both the raw and estimated pedestrian movement data when combined with the aggregated crash data. This investigation shows the potential utility of such a model that is both scalable to larger geographic areas, and translatable to varying jurisdictions due to its reliance on nationally-available datasets. Policy implications and concerns surrounding use of the Safety In Numbers effect in planning and engineering, and issues of data quality and availability in urban geographic science, are discussed.
  • Thumbnail Image
    Item
    Data, R Code, and Output Supporting "An Historical Overview and Update of Wolf-Moose Interactions in Northeastern Minnesota"
    (2017-10-06) Fieberg, John R; Mech, L. David; Barber-Meyer, Shannon; jfieberg@umn.edu; Fieberg, John R
    These files contain data and R code (along with associated output from running the code) supporting all results reported in, "Mech, L. D., J. Fieberg, and S. Barber-Meyer. In press. An historical overview and update of wolf-moose interactions in Northeastern Minnesota. Wildlife Society Bulletin." In this paper, we explored relationships between wolf numbers, monitored in part of the Minnesota moose range, and moose calf:population and estimated log annual growth rates of moose in Northeast Minnesota.
  • Thumbnail Image
    Item
    Modeling Outputs of Efficient Compressibility Estimators
    (2018-06) Asamoah Owusu, Dennis
    There are times when it is helpful to know whether data is compressible before expending computational resources to compress it. The standard deviation of the byte distribution of data is an example of a measure of compressibility that does not involve actually compressing the data. This work considered five such measures of compressibility: byte standard deviation, shannon entropy, “average meaning entropy”, “byte counting” and “heuristic method”. We developed models that relate the output of these measures to the compression ratios of gzip, lz4 and xz using data retrieved from browsing Facebook, Wikipedia and YouTube. The models for byte standard deviation, shannon entropy and “average meaning entropy” were linear in both the parameters and the variables. The model for “byte counting” was non-linear in the predictor variable but linear in the parameters. The “heuristic method” was a classification model. In general, there was a strong relationship between the measures and the compressibility of a given data. Also, in many cases the models developed using one set of data from a source (like Youtube) was able to estimate the compressibility of another data set from the same source to a useful extent. This suggests the potential for developing a model per ECE for a source that can predict, to a useful degree, the compressibility of data from that source. At the same time, the differences in accuracy when models were evaluated on the data they were developed from versus when evaluated on new data from the same source indicate that there are important differences in the nature of the data coming from even the same source.