This repository contains the code reproducing the results in the paper "Active Learning for Abstract Models of Collectives", by Alexander Schiendorfer, Christoph Lassner, Gerrit Anders, Wolfgang Reif and Rainer Lienhart.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
300plants
400plants
analysisUtils
ALEQ-comp-init-20-400-plants.pdf
ALEQ-comp-init-30-400-plants.pdf
Costs50.csv
PLF.py
README.md
comparison_GPs_DFs.ipynb
integrated_active_learning_eval.ipynb
tradeoff400.pdf

README.md

active_learning_for_abstract_models_of_collectives

This repository contains the code reproducing the results in the papers "Active Learning for Abstract Models of Collectives" [1] and "Active Learning for Efficient Sampling of Control Models of Collectives" [2], by Alexander Schiendorfer, Christoph Lassner, Gerrit Anders, Wolfgang Reif and Rainer Lienhart.

a) comparison_GPs_DFs.ipynb presents the results of two different active learning strategies using Gaussian Processes (GPs) and Decision Forests (DFs) on a single AVPP dataset (costs50.csv) shown in [1] and [2] b) integrated_active_learning_eval.ipynb presents the analysis of the effectiveness of DF-AL integrated into a full simulation environment [2]

The full source code of the simulation environment used to produce the results in b) is found at https://github.com/Alexander-Schiendorfer/active-learning-collectives .