Pump Prime
One pioneering approach that showed enormous success is the Pump Priming Programme. While providing money for travel and secondments is always effective, there is nothing like the opportunity for research funding to focus the minds of academics. Knowledge 4 All continues its role of a mini funding agency and encouraged its members and others outside the field to collaborate and perform exciting new research together by supporting a series of preliminary investigations in areas of interest to the network.Past projects
Learning and Inference in Structured, Higher Dimensional Stochastic Differential Systems
Learning and Inference in Structured, Higher Dimensional Stochastic Differential Systems
Modelling Learning and Co-Learning Curves for Application to Cognitive-Interfaces
Studying learning dynamics in multi-component systems, focusing on optimizing human-machine interactions through adaptive interfaces
Sparse Reinforcement Learning in High Dimensions 1
Creating efficient reinforcement learning algorithms for high-dimensional data to improve decision-making
Semi-supervised learning of semantic spatial concepts for a mobile robot
Designing algorithms for robots to learn and adapt to environments with minimal human intervention
Context Models for Textual Entailment and their Application to Statistical Machine Translation
Enhancing machine translation with probabilistic context models and semantic inference.
Data-Dependent Geometries and Structures : Analyses and Algorithms for Machine Learning
Exploring how data-driven geometries can enhance machine learning analysis and algorithms
Online Learning for Robot Control
Applying online learning methods to improve robot controllers for smooth and precise operations
Bridging machine classification and human cognition: discriminative vs. generative learning
Connecting machine learning methods with human cognitive approaches to classification
Sparse Reinforcement Learning in High Dimensions 2
Advancing reinforcement learning for high-dimensional domains using compressive sensing techniques
Normalised Compression Distance Measures and Their Application in Unsupervised and Supervised Analysis of Polymorphic Data
Using compression-based methods for analyzing and classifying complex datasets
Learning with Labeled and Unlabeled Data
Leveraging unlabeled data to improve semi-supervised learning for scalable and efficient applications
Dynamics and Uncertainty in Interaction
Innovating interaction methods for brain-computer interfaces and exploratory navigation of musical spaces
Large Margin Algorithms and Kernel Methods for Speech Recognition
Developing robust kernel-based algorithms for improving speech phoneme classification
Multi-Task Learning: Optimisation Methods and Applications
Creating multi-task learning algorithms with applications in marketing, bioinformatics, and robotics
CARTER
Investigating scene categorization using document analysis algorithms for improved image retrieval and annotation
Next Generation Information Retrieval
Integrating semantic information to improve cross-lingual information retrieval using Wikipedia as a resource
GISK
Exploring grammatical inference algorithms for efficiently learning context-sensitive languages
Methods for fusing eye movements and text content for information retrieval
ombining eye-tracking data and text to predict document relevance in information retrieval
Online Performance of Reinforcement Learning with Internal Reward Functions
Focusing on reinforcement learning with online paradigms to improve intelligent interfaces
