oKDE-Java

This is a Java implementation of the oKDE algorithm proposed by Matej Kristan (oKDE). This project is part of my master's thesis that uses the oKDE algorithm to estimate a model of human mobility and exploits the estimated model to predict future locations of human individuals (Location Prediction Based on Mobility Patterns in Location Histories: git repository).

---

1. oKDE Algorithm
2. Build Instructions
3. Quickstart
4. External Libraries Used

---

## oKDE Algorithm

Given a set of n-dimensional samples, this algorithm estimates the distribution of the samples using kernel density estimation. The output model is a Mixture of Gaussians. Moreover, the model is compressed and can be updated as new samples arrive.

The basic principle of the oKDE algorithm is summarized by this graphic:

For more details see (oKDE).

## Build Instructions

Just execute ant in project root to compile the project:

$ ant

Afterwards, the packed jar file can be found in the dist-folder.

## Quickstart

Here you can find a simple example that uses oKDE-Java to estimate a distribution of randomly generated samples. This example illustrates the basic usage of oKDE-java.

To use oKDE-Java in another project just include the jar file (see above how to build).

## External Libraries Used

The following libraries are used in oKDE-Java:

• EJML, a linear algebra library for manipulating dense matrices License: http://www.apache.org/licenses/LICENSE-2.0