Incremental Online Learning in High Dimensions

By Sethu Vijayakumar, Aaron D'Souza, Stefan Schaal. 2005

613 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Mechanics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Mechanics

In Computer Vision and Pattern Recognition

Citations in the corpus

BM: An iterative algorithm to learn stable non-linear dynamical systems with Gaussian mixture models (2010, methodology): “LWPR in comparison offers a cost-efficient method for non-linear function approximation; see [12] for a comparison.”
Surface EMG in advanced hand prosthetics (2008, methodology): “The simple idea behind batch uniformisation is that, in a reallife set-up such as ours, there can be many input samples located in the very same region of the input space, with very similar target values.”
A probabilistic approach to mixed open-loop and closed-loop control, with application to extreme autonomous driving (2010, background): “However, with few exceptions the term “multiple models” is typically synonymous with “local models,” building or learning a collection of models that are each accurate in a different portion of the state space – for example, locally linear models, [18], [5], [24] or nonparametric models such as Gaus”
Learning bimanual end-effector poses from demonstrations using task-parameterized dynamical systems (2015, background): “Several learning algorithms have been proposed to encode human demonstrations in a compact way, while encapsulating the relevant information of the task at hand [3], [4], [5].”
Local Gaussian Process Regression for Real Time Online Model Learning (2008, methodology): “These problems have been addressed by real-time learning methods such as locally weighted projection regression (LWPR) [7,8].”
Incremental learning with self-organizing maps (2017, background): “Although attempts at formal definition ([2], [1]) conflict in minor points, they seem to agree that incremental learning algorithms”
Learning Inverse Dynamics by Gaussian process Begrression under the Multi-Task Learning Framework (2009, methodology): “The method known as locally weighted projection regression (LWPR) [21, 20, 19 ] is currently the standard learning method used in the robotics community since it is capable of online, real-time learning even for complex robots such as humanoid robots.”
Learning Physical Collaborative Robot Behaviors From Human Demonstrations (2016, methodology): “In contrast with other regression methods such as locally weighted projection regression [47] or Gaussian”
Sensor-assisted adaptive motor control under continuously varying context (2007, methodology): “For learning the mapping τ(θ, θ̇, θ̈, s), regression techniques exist, like locally-weighted projection re- gression (Vijayakumar et al., 2005), that can exploit this reduced intrinsic dimensionality efficiently.”
Learning control Lyapunov function to ensure stability of dynamical system-based robot reaching motions (2014, background): “GMR is computationally fast, but is comparatively less accurate, for instance, than GPR. Locally Weighted Projection Regression (LWPR) [16] is a powerful tool for incremental learning but requires setting several initial parameters.”