Tom Diethe

Constructing nonlinear discriminants from multiple data views

There are many situations in which we have more than one view of a single data source, or in which we have multiple sources of data that are aligned. We would like to be able to build classifiers which incorporate these to enhance classification performance. Kernel Fisher Discriminant Analysis (KFDA) can be formulated as a convex optimisation problem, which we extend to the Multiview setting (MFDA) and introduce a sparse version (SMFDA). We show that our formulations are justified from both probabilistic and learning theory perspectives. We then extend the optimisation problem to account for directions unique to each view (PMFDA). We show experimental validation on a toy dataset, and then give experimental results on a brain imaging dataset and part of the PASCAL 2007 VOC challenge dataset.

Compressed Sampling for pulse Doppler radar

This paper presents a study of how the Analogue to Digital Converter (ADC) sampling rate in a digital radar can be reduced-without reduction in waveform bandwidth-through the use of Compressed Sampling (CS). Real radar data is used to show that through use of chirp or Gabor dictionaries and Basis Pursuit (BP) the ADC sampling frequency can be reduced by a factor of 128, to under 1 mega sample per second, while the waveform bandwidth remains 40 MHz. The error on the reconstructed fast-time samples is small enough that accurate range-profiles and range-frequency surfaces can be produced.

Online learning with multiple kernels: A review

This review examines kernel methods for online learning, in particular, multiclass classification. We examine margin-based approaches, stemming from Rosenblatts original perceptron algorithm, as well as nonparametric probabilistic approaches that are based on the popular gaussian process framework. We also examine approaches to online learning that use combinations of kernels—online multiple kernel learning. We present empirical validation of a wide range of methods on a protein fold recognition data set, where different biological feature types are available, and two object recognition data sets, Caltech101 and Caltech256, where multiple feature spaces are available in terms of different image feature extraction methods.

SPHERE: A Sensor Platform for Healthcare in a Residential Environment

It can be tempting to think about smart homes like one thinks about smart cities. On the surface, smart homes and smart cities comprise coherent systems enabled by similar sensing and interactive technologies. It can also be argued that both are broadly underpinned by shared goals of sustainable development, inclusive user engagement and improved service delivery. However, the home possesses unique characteristics that must be considered in order to develop effective smart home systems that are adopted in the real world [37].

Unsupervised learning of sensor topologies for improving activity recognition in smart environments

There has been significant recent interest in sensing systems and smart environments, with a number of longitudinal studies in this area. Typically the goal of these studies is to develop methods to predict, at any one moment of time, the activity or activities that the resident(s) of the home are engaged in, which may in turn be used for determining normal or abnormal patterns of behaviour (e.g. in a health-care setting). Classification algorithms, such as Conditional Random Field (CRFs), typically consider sensor activations as features but these are often treated as if they were independent, which in general they are not. Our hypothesis is that learning patterns based on combinations of sensors will be more powerful than single sensors alone. The exhaustive approach to take all possible combinations of sensors and learn classifier weights for each combination is clearly computationally prohibitive. We show that through the application of signal processing and information-theoretic techniques we can learn about the sensor topology in the home (i.e. learn an adjacency matrix) which enables us to determine the combinations of sensors that will be useful for classification ahead of time. As a result we can achieve classification performance better than that of the exhaustive approach, whilst only incurring a small cost in terms of computational resources. We demonstrate our results on several datasets, showing that our method is robust in terms of variations in the layout and the number of residents in the house. Furthermore, we have incorporated the adjacency matrix into the CRF learning framework and have shown that it can improve performance over multiple baselines.

An RSSI-based wall prediction model for residential floor map construction

In residential environments, floor maps, often required by location-based services, cannot be trivially acquired. Researchers have addressed the problem of automatic floor map construction in indoor environments using various modalities, such as inertial sensors, Radio Frequency (RF) fingerprinting and video cameras. Considering that some of these techniques are unavailable or impractical to implement in residential environments, in this paper, we focus on using RF signals to predict the number of walls between a wearable device and an access point. Using both supervised and unsupervised learning techniques on two data sets; a system-level data set of Bluetooth packets, and measurements on the signal attenuation, we construct wall prediction models that yield up to 91% identification rate. As a proof-of-concept, we also use the wall prediction models to infer the floor plan of a smart home deployment in a real residential environment.

A Comprehensive Study of Activity Recognition Using Accelerometers

This paper serves as a survey and empirical evaluation of the state-of-the-art in activity recognition methods using accelerometers. The paper is particularly focused on long-term activity recognition in real-world settings. In these environments, data collection is not a trivial matter; thus, there are performance trade-offs between prediction accuracy, which is not the sole system objective, and keeping the maintenance overhead at minimum levels. We examine research that has focused on the selection of activities, the features that are extracted from the accelerometer data, the segmentation of the time-series data, the locations of accelerometers, the selection and configuration trade-offs, the test/retest reliability, and the generalisation performance. Furthermore, we study these questions from an experimental platform and show, somewhat surprisingly, that many disparate experimental configurations yield comparable predictive performance on testing data. Our understanding of these results is that the experimental setup directly and indirectly defines a pathway for context to be delivered to the classifier, and that, in some settings, certain configurations are more optimal than alternatives. We conclude by identifying how the main results of this work can be used in practice, specifically in experimental configurations in challenging experimental conditions.

Kernel Polytope Faces Pursuit

Polytope Faces Pursuit (PFP) is a greedy algorithm that approximates the sparse solutions recovered by ***1 regularised least-squares (Lasso) [4,10] in a similar vein to (Orthogonal) Matching Pursuit (OMP) [16]. The algorithm is based on the geometry of the polar polytope where at each step a basis function is chosen by finding the maximal vertex using a path-following method. The algorithmic complexity is of a similar order to OMP whilst being able to solve problems known to be hard for (O)MP. Matching Pursuit was extended to build kernel-based solutions to machine learning problems, resulting in the sparse regression algorithm, Kernel Matching Pursuit (KMP) [17]. We develop a new algorithm to build sparse kernel-based solutions using PFP, which we call Kernel Polytope Faces Pursuit (KPFP). We show the usefulness of this algorithm by providing a generalisation error bound [7] that takes into account a natural regression loss and experimental results on several benchmark datasets.

Releasing eHealth Analytics into the Wild: Lessons Learnt from the SPHERE Project

The SPHERE project is devoted to advancing eHealth in a smart-home context, and supports full-scale sensing and data analysis to enable a generic healthcare service. We describe, from a data-science perspective, our experience of taking the system out of the laboratory into more than thirty homes in Bristol, UK. We describe the infrastructure and processes that had to be developed along the way, describe how we train and deploy Machine Learning systems in this context, and give a realistic appraisal of the state of the deployed systems.

BDL.NET: Bayesian dictionary learning in Infer.NET

We introduce and analyse a flexible and efficient implementation of Bayesian dictionary learning for sparse coding. By placing Gaussian-inverse-Gamma hierarchical priors on the coefficients, the model can automatically determine the required sparsity level for good reconstructions, whilst also automatically learning the noise level in the data, obviating the need for heuristic methods for choosing sparsity levels. This model can be solved efficiently using Variational Message Passing (VMP), which we have implemented in the Infer.NET framework for probabilistic programming and inference. We analyse the properties of the model via empirical validation on several accelerometer datasets. We provide source code to replicate all of the experiments in this paper.