Animations
Airport Algorithm
![]() |
|
|
Multi-Value-Functions: Efficient Automatic Action Hierarchies for Multiple Goal MDPs |
|
|
Quickly computing all policies for all goals in a noisy maze |
movie |
|
Example execution of "all goals" policy |
|
|
Using "all goals" to chase stupid bad guys |
|
|
Using "all goals" to chase simplistic bad guys |
|
|
Using "all goals" to chase frightened bad guys |
|
|
Using "all goals" to chase intelligent bad guys |
Anchors
![]() |
|
|
The Anchors Hierarchy: Using the Triangle Inequality to Survive High-Dimensional Data |
paper |
|
Efficient spatial partioning with size criterion |
movie |
|
Efficient spatial partioning with space criterion |
Anomaly Hunting
![]() |
|
|
Using Tarjan's Red Rule for Fast Dependency Tree Construction |
paper |
|
Finding anomalous records in huge datasets using human feedback to focus on interesting galaxies in Sloan data |
movie |
Robot Skill Learning
|
|
|
High Dimension Action Spaces in Robot Skill Learning |
paper |
|
A PUMA robot learns to throw a tennis ball in 100 trials using combination of active learning and kd-trees ( learning phase ) |
movie |
FLINK
![]() |
|
|
Stochastic Link and Group Detection |
|
|
Learning social networks from massive noisy transactional data |
movie |
GoodWalk
|
|
|
Policy Search by Dynamic Programming |
paper |
|
Fast search for good walking policies on [Zeglin03] robot simulator using logistic regression and PSDP |
movie |
Kmeans
![]() |
|
|
Accelerating Exact k-means Algorithms with Geometric Reasoning |
|
|
Accelerated learning of K-means model and automatic estimation of number of clusters |
movie |
Dualtree K-Nearest Neighbor
![]() |
|
|
N-Body Problems in Statistical Learning |
paper |
|
Fast search for the k-nearest neighbor data points to some query points in k-dimensional space (different k!) |
movie |
Fast Mixture Learning
![]() |
|
|
Very Fast EM-based Mixture Model Clustering Using Multiresolution KD-trees |
paper |
|
Accelerating Gaussian Mixture Learning with Multi-resolution trees |
movie |
PARTI-GAME
![]() |
|
|
The Parti-game Algorithm for Variable Resolution Reinforcement Learning in Multidimensional State-spaces |
paper |
|
Multi-resolution reinforcement learning algorithm |
movie |
Pool playing
![]() |
|
|
Locally Weighted Learning for Control |
|
|
Robot learning to play pool using locally weighted linear regression |
movie |
Fast Scan Statistics
![]() |
|
|
Rapid Detection of Significant Spatial Clusters |
|
|
Analysing National Retail Pharmacy Sales data to find regions with unusual increase in counts |
movie |
Variable Resolution Dynamic Programming
![]() |
|
|
Variable Resolution Discretization in Optimal Control |
paper |
|
Optimal control of acrobot (two joint arm with unpowered shoulder) using automatically learned partioning of state space |
movie |
|
Optimal control of car-on-hill using automatically learned partioning of state space |
movie |










