Trend-Based Prediction of Spatial Change

Authors: Xiaoyu Ge, Jae Hee Lee, Jochen Renz, Peng Zhang

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our method using real world data from two completely different domains, which demonstrates the accuracy of our predictions. We applied the method to two different spatial domains, namely wild fire progression and also cloud movement, where it is more likely that sudden changes or topological changes of a region occur. We use a real-world data set for each domain and the average precision and recall under different settings are summarized in Table. 2.
Researcher Affiliation Academia Research School of Computer Science, Australian National University {xiaoyu.ge,jochen.renz,p.zhang}@anu.edu.au QCIS, FEIT, University of Technology Sydney jaehee.lee@uts.edu.au
Pseudocode Yes Algorithm 1: Boundary point prediction
Open Source Code No The paper does not provide a link to open-source code or state that the code is publicly available.
Open Datasets Yes Wild Fire Progression: From the data base of USDA forest service2, we obtained a sequence of progression maps of a real wild fire. 2http://www.fs.fed.us/nwacfire/ball/. Cloud Movement We also obtained video data of moving clouds that is recorded using a fisheye lens (see [Wood-Bradley et al., 2012] for detail).
Dataset Splits No The paper describes using past snapshots for prediction and evaluates on real-world and generated data, but it does not specify explicit training/validation/test splits (e.g., percentages or sample counts) or cross-validation setup.
Hardware Specification No The paper does not specify any particular hardware (e.g., CPU, GPU models, or cloud instance types) used for running the experiments.
Software Dependencies No The paper does not provide specific software names with version numbers, such as libraries or solvers.
Experiment Setup Yes For each evaluation, our method takes a sequence RT w+1, . . . , RT of w recent snapshots to predict the outer approximation of regions at T + 1. For our analysis we varied both w and the threshold d as defined in Section 4.3.