
Switched linear projections and inactive state sensitivity for deep neural network interpretability
We introduce switched linear projections for expressing the activity of ...
read it

Document Clustering Evaluation: Divergence from a Random Baseline
Divergence from a random baseline is a technique for the evaluation of d...
read it

AutoJacoBin: Autoencoder Jacobian Binary Hashing
Binary codes can be used to speed up nearest neighbor search tasks in la...
read it

Practical WholeSystem Provenance Capture
Data provenance describes how data came to be in its present form. It in...
read it

PseudoRecursal: Solving the Catastrophic Forgetting Problem in Deep Neural Networks
In general, neural networks are not currently capable of learning tasks ...
read it

Some Approximation Bounds for Deep Networks
In this paper we introduce new bounds on the approximation of functions ...
read it

The effect of the choice of neural network depth and breadth on the size of its hypothesis space
We show that the number of unique function mappings in a neural network ...
read it

Adaptive Approximation Error Models for Efficient Uncertainty Quantification with Application to Multiphase Subsurface Fluid Flow
Samplebased Bayesian inference provides a route to uncertainty quantifi...
read it

Negative type diversities, a multidimensional analogue of negative type metrics
Diversities are a generalization of metric spaces in which a nonnegativ...
read it

An O(n n) time Algorithm for computing the Pathlength Distance between Trees
Tree comparison metrics have proven to be an invaluable aide in the reco...
read it

PseudoRehearsal: Achieving Deep Reinforcement Learning without Catastrophic Forgetting
Neural networks can achieve extraordinary results on a wide variety of t...
read it

On Enactability and Monitorability of Agent Interaction Protocols
Interactions between agents are usually designed from a global viewpoint...
read it

On the Enactability of Agent Interaction Protocols: Toward a Unified Approach
Interactions between agents are usually designed from a global viewpoint...
read it

Unsupervised Domain Adaptation using Deep Networks with CrossGrafted Stacks
Popular deep domain adaptation methods have mainly focused on learning d...
read it

Incorporating social practices in BDI agent systems
When agents interact with humans, either through embodied agents or beca...
read it

Continuous Integration Theater
Background: Continuous Integration (CI) systems are now the bedrock of s...
read it

Cyclic Sequence Generators as Program Counters for HighSpeed FPGAbased Processors
This paper compares the performance of conventional radix2 program coun...
read it

Bayesian inference of species trees using diffusion models
We describe a new and computationally efficient Bayesian methodology for...
read it

GRImRePR: Prioritising Generating Important Features for PseudoRehearsal
Pseudorehearsal allows neural networks to learn a sequence of tasks wit...
read it

MIME: Mutual Information Minimisation Exploration
We show that reinforcement learning agents that learn by surprise (surpr...
read it
University of Otago
The University of Otago is New Zealand's first university and a vibrant international centre of learning. It was established in 1869 by Scottish settlers with a strong conviction in the transforming power of education. Today the University has abou...