A probabilistic model with parsinomious representation for sensor fusion in recognizing activity in pervasive environment

To tackle the problem of increasing numbers of state transition parameters when the number of sensors increases, we present a probabilistic model together with several parsinomious representations for sensor fusion. These include context specific independence (CSI), mixtures of smaller multinomials and softmax function representations to compactly represent the state transitions of a large number of sensors. The model is evaluated on real-world data acquired through ubiquitous sensors in recognizing daily morning activities. The results show that the combination of CSI and mixtures of smaller multinomials achieves comparable performance with much fewer parameters.

Recognising behaviours of multiple people with hierarchical probabilistic model and statistical data association

Recognising behaviours of multiple people, especially high-level behaviours, is an important task in surveillance systems. When the reliable assignment of people to the set of observations is unavailable, this task becomes complicated. To solve this task, we present an approach, in which the hierarchical hidden Markov model (HHMM) is used for modeling the behaviour of each person and the joint probabilistic data association filters (JPDAF) is applied for data association. The main contributions of this paper lie in the integration of multiple HHMMs for recognising high-level behaviours of multiple people and the construction of the Rao-Blackwellised particle filters (RBPF) for approximate inference. Preliminary experimental results in a real environment show the robustness of our integrated method in behaviour recognition and its advantage over the use of Kalman filter in tracking people.

Preference networks : probabilistic models for recommendation systems

Recommender systems are important to help users select relevant and personalised information over massive amounts of data available. We propose an unified framework called Preference Network (PN) that jointly models various types of domain knowledge for the task of recommendation. The PN is a probabilistic model that systematically combines both content-based filtering and collaborative filtering into a single conditional Markov random field. Once estimated, it serves as a probabilistic database that supports various useful queries such as rating prediction and top-N recommendation. To handle the challenging problem of learning large networks of users and items, we employ a simple but effective pseudo-likelihood with regularisation. Experiments on the movie rating data demonstrate the merits of the PN.

Preference Networks: probabilistic models for recommendation systems

Abstract
Recommender systems are important to help users select relevant and personalised information over massive amounts of data available. We propose an unified framework called Preference Network (PN) that jointly models various types of domain knowledge for the task of recommendation. The PN is a probabilistic model that systematically combines both content-based filtering and collaborative filtering into a single conditional
Markov random field. Once estimated, it serves as a probabilistic database that supports various useful queries such as rating prediction and top-N recommendation. To handle the challenging problem of learning large networks of users and items, we employ a simple but effective pseudo-likelihood with regularisation. Experiments on the movie rating data demonstrate the merits of the PN.

Adaptive load shedding for mining frequent patterns from data streams

Most algorithms that focus on discovering frequent patterns from data streams assumed that the machinery is capable of managing all the incoming transactions without any delay; or without the need to drop transactions. However, this assumption is often impractical due to the inherent characteristics of data stream environments. Especially under high load conditions, there is often a shortage of system resources to process the incoming transactions. This causes unwanted latencies that in turn, affects the applicability of the data mining models produced – which often has a small window of opportunity. We propose a load shedding algorithm to address this issue. The algorithm adaptively detects overload situations and drops transactions from data streams using a probabilistic model. We tested our algorithm on both synthetic and real-life datasets to verify the feasibility of our algorithm.

Sculpting fog: attempting to define "acceptable risk" in emergency services.

There is currently no consensus as to how “acceptable risk” should be defined in emergency service response. Attempts to address this have relied upon the assumption that a probabilistic model of risk can be calculated and that acceptable levels of risk can be determined. Examples of this process can be seen in a number of emergency services, e.g. dynamic risk assessment utilised by a number of fire services.

Multivariate modelling of variability in sheet metal forming

The inherent variability in incoming material and process conditions in sheet metal forming makes quality control and the maintenance of consistency extremely difficult. A single FEM simulation is successful at predicting the formability for a given system, however lacks the ability to capture the variability in an actual production process due to the numerical deterministic nature. This paper investigates a probabilistic analytical model where the variation of five input parameters and their relationship to the sensitivity of springback in a stamping process is examined. A range of sheet tensions are investigated, simulating different operating windows in an attempt to highlight robust regions where the distribution of springback is small. A series of FEM simulations were also performed, to compare with the findings from the analytical model using AutoForm Sigma v4.04 and to validate the analytical model assumptions.

Results show that an increase in sheet tension not only decreases springback, but more importantly reduces the sensitivity of the process to variation. A relative sensitivity analysis has been performed where the most influential parameters and the changes in sensitivity at various sheet tensions have been investigated. Variation in the material parameters, yield stress and n-value were the most influential causes of springback variation, when compared to process input parameters such as friction, which had a small effect. The probabilistic model presented allows manufacturers to develop a more comprehensive assessment of the success of their forming processes by capturing the effects of inherent variation.

Facial age estimation by nonlinear aging pattern subspace

Human age estimation by face images is an interesting yet challenging research topic emerging in recent years. This paper extends our previous work on facial age estimation (a linear method named AGES). In order to match the nonlinear nature of the human aging progress, a new algorithm named KAGES is proposed based on a nonlinear subspace trained on the aging patterns, which are defined as sequences of individual face images sorted in time order. Both the training and test (age estimation) processes of KAGES rely on a probabilistic model of KPCA. In the experimental results, the performance of KAGES is not only better than all the compared algorithms, but also better than the human observers in age estimation. The results are sensitive to parameter choice however, and future research challenges are identified.

Ordinal Boltzmann machines for collaborative filtering

Collaborative filtering is an effective recommendation technique wherein the preference of an individual can potentially be predicted based on preferences of other members. Early algorithms often relied on the strong locality in the preference data, that is, it is enough to predict preference of a user on a particular item based on a small subset of other users with similar tastes or of other items with similar properties. More recently, dimensionality reduction techniques have proved to be equally competitive, and these are based on the co-occurrence patterns rather than locality. This paper explores and extends a probabilistic model known as Boltzmann Machine for collaborative filtering tasks. It seamlessly integrates both the similarity and cooccurrence in a principled manner. In particular, we study parameterisation options to deal with the ordinal nature of the preferences, and propose a joint modelling of both the user-based and item-based processes. Experiments on moderate and large-scale movie recommendation show that our framework rivals existing well-known methods.

Application of computational media aesthetics methodology to extracting color semantics in film

Using film grammar as the underpinning, we study the extraction of structures in video based on color using a wide configuration of clustering methods combined with existing and new similarity measures. We study the visualisation of these structures, which we call Scene-Cluster Temporal Charts and show how it can bring out the interweaving of different themes and settings in a film. We also extract color events that filmmakers use to draw/force a viewer's attention to a shot/scene. This is done by first extracting a set of colors used rarely in film, and then building a probabilistic model for color event detection. We demonstrate with experimental results from ten movies that our algorithms are effective in the extraction of both scene-cluster temporal charts and color events.

Learning Boltzmann distance metric for face recognition

We introduce a new method for face recognition using a versatile probabilistic model known as Restricted Boltzmann Machine (RBM). In particular, we propose to regularise the standard data likelihood learning with an information-theoretic distance metric defined on intra-personal images. This results in an effective face representation which captures the regularities in the face space and minimises the intra-personal variations. In addition, our method allows easy incorporation of multiple feature sets with controllable level of sparsity. Our experiments on a high variation dataset show that the proposed method is competitive against other metric learning rivals. We also investigated the RBM method under a variety of settings, including fusing facial parts and utilising localised feature detectors under varying resolutions. In particular, the accuracy is boosted from 71.8% with the standard whole-face pixels to 99.2% with combination of facial parts, localised feature extractors and appropriate resolutions.

A probabilistic movement model for shortest path formation in virtual ant-like agents

We propose a probabilistic movement model for controlling ant-like agents foraging between two points. Such agents are all identical, simple, autonomous and can only communicate indirectly through the environment. These agents secrete two types of pheromone, one to mark trails towards the goal and another to mark trails back to the starting point. Three pheromone perception strategies are proposed (Strategy A, B and C). Agents that use strategy A perceive the desirability of a neighbouring location as the difference between levels of attractive and repulsive pheromone in that location. With strategy B, agents perceive the desirability of a location as the quotient of levels of attractive and repulsive pheromone. Agents using strategy C determine the product of the levels of attractive pheromone with the complement of levels of repulsive pheromone. We conduct experiments to confirm directionality as emergent property of trails formed by agents that use each strategy. In addition, we compare path formation speed and the quality of the formed path under changes in the environment. We also investigate each strategy's robustness in environments that contain obstacles. Finally, we investigate how adaptive each strategy is when obstacles are eventually removed from the scene and find that the best strategy of these three is strategy A. Such a strategy provides useful guidelines to researchers in further applications of swarm intelligence metaphors for complex problem solving.