Termine

Aktuelles

UNIGIS Abschlussarbeiten


Der krönende Abschluss eines UNIGIS MSc Studiums ist sicherlich die Master Thesis. Mit ihr belegen unsere MSc-AbsolventInnen, dass sie den akademischen Grad "Master of Science (Geographical Information Science & Systems)" zu Recht führen.  Im UNIGIS professional Studiengang muss keine Abschlussarbeit verfasst werden. Dennoch nehmen einige Studierende die Möglichkeit war, ein Geoinformatikprojekt durchzuführen und entsprechend zu dokumentieren.

Sie sind auf der Suche nach aktueller Literatur zu Geoinformatik-Themen?
Hier finden sie die mitunter preisgekrönten Abschlussarbeiten unserer AbsolventInnen!

Nachricht
  • Failed loading XML...

Achim Guenter Hettel [01-2015]:

Prediction Accuracy of Hot Spot Methods

Diese Arbeit ist online verfügbar: Download


Prediction Accuracy of Hot Spot Methods Crime hot spot mapping has gained increased interest during past several years among researchers and crime analysts. Especially the subject of forecasting where crime tends to occur has been focused on. Many different techniques have been developed to identify patterns of crime to support researchers and analysts to examine where crime may occur in the future by the use of retrospective crime patterns. So far, no research included offender residence patterns in its calculation. This thesis focused on the prediction accuracy of different hot spot methods based on just one crime type and the related offender residences in the city of Anchorage, USA. The main goal was to find a cluster method which outperforms all other methods based on its prediction accuracy related to three different retrospective time periods. Additionally, the crime type and offender residence data were compared based on the outcome of the applied calculation and the influences of limited prospective point numbers were examined. The first step in this research included the examination of the global distribution of both data patterns from all time periods. Nine different cluster techniques were applied to visually detect hot spots of crime events and offender residences. The related hot spots were then spatially analyzed and the results included into three prediction accuracy calculations. Most previously conducted research on a similar topic concluded that the kernel density estimation method (KDE) has the best overall prediction accuracy performance. The results of this thesis confirmed that the KDE method outperforms all other methods in most of the different time periods containing the crime type. However, the grid based Gi* method showed the best prediction accuracy performances related to the offender residence data pattern. Comparing the two data sets, the crime event data set showed the highest prediction accuracy indexes on average. Another finding was that the influence of limited numbers of prospective data was causing several hot spot methods to fail calculating the prediction indexes. By comparing all best indexes throughout the time periods and data sets, the KDE method exceeded all other methods.


Zurck weitere Arbeiten ...

Club UNIGIS Login

Mitglieder der Club-UNIGIS Absolventennetzwerkes melden sich hier mit Ihren Blackboard-Zugangsdaten an: