In this project we investigate the use of artificial neural networks(ANNs) as the core prediction function of a recommender system. In the past, research concerned with recommender systems that use ANNs have mainly concentrated on using collaborative-based information. We look at the effects of adding content-based information and how altering the topology of the network itself affects the accuracy of the recommendations generated. In particular, we investigate a mixture of experts topology. We create two expert clusters in the hidden layer of the ANN, one for content-based data and another for collaborative-based data. This greatly reduces the number of connections between the input and hidden layers. Our experimental evaluation shows that this new architecture produces the same accuracy of recommendation as the fully connected configuration with a large decrease in the amount of time it takes to train the network. This decrease in time is a great advantage because of the need for recommender systems to provide real time results to the user.

This paper proposes a mixture of experts recurrent events model. This general model accommodates an unobservable frailty variable, intervention effect, influence of accumulating event occurrences, and covariate effects. A latent class variable is utilized to deal with a heterogeneous population and associated covariates. A homogeneous nonparametric baseline hazard and heterogeneous parametric covariate effects are assumed. Maximum likelihood principle is employed to obtain parameter estimates. Since the frailty variable and latent classes are unobserved, an estimation procedure is derived through the EM algorithm. A simulated data set is generated to illustrate the data structure of recurrent events for a heterogeneous population.

This thesis investigates using adaptive estimation techniques to determine unknown model parameters such as size and surface material reflectivity, while estimating position, velocity, attitude, and attitude rates of a resident space object. This work focuses on the application of these methods to the space situational awareness problem. This thesis proposes a unique method of implementing a top-level gating network in a dual-layer hierarchical mixture of experts. In addition it proposes a decaying learning parameter for use in both the single layer mixture of experts and the dual-layer hierarchical mixture of experts. Both a single layer mixture of experts and dual-layer hierarchical mixture of experts are compared to the multiple model adaptive estimation in estimating resident space object parameters such as size and reflectivity. The hierarchical mixture of experts consists of macromodes. Each macromode can estimate a different parameter in parallel. Each macromode is a single layer mixture of experts with unscented Kalman filters used as the experts. A gating network in each macromode determines a gating weight which is used as a hypothesis tester. Then the output of the macromode gating weights go to a top level gating weight to determine which macromode contains the most probable model. The measurements consist of astrometric and photometric data from non-resolved observations of the target gathered via a telescope with a charge coupled device camera. Each filter receives the same measurement sequence. The apparent magnitude measurement model consists of the Ashikhmin Shirley bidirectional reflectance distribution function. The measurements, process models, and the additional shape, mass, and inertia characteristics allow the algorithm to predict the state and select the most probable fit to the size and reflectance characteristics based on the statistics of the measurement residuals and innovation covariance. A simulation code is developed to test these adaptive estimation techniques. The feasibility of these methods will be demonstrated in this thesis.

Access technologies often deal with the classification of several physiological and biomechanical signals. In most previous studies involving access technologies, a single classifier has been trained. Despite reported success of these single classifiers, classification accuracies are often below clinically viable levels. One approach to improve upon the performance of these classifiers is to utilize the state of- the-art multiple classifier systems (MCS). Because MCS invoke more than one classifier, more information can be exploited from the signals, potentially leading to higher classification performance than that achievable with single classifiers. Moreover, by decreasing the feature space dimensionality of each classifier, the speed of the system can be increased. MCSs may combine classifiers on three levels: abstract, rank, or measurement level. Among them, abstract-level MCSs have been the most widely applied in the literature given the flexibility of the abstract level output, i.e., class labels may be derived from any type of classifier and outputs from multiple classifiers, each designed within a different context, can be easily combined. In this thesis, we develop two new abstract-level MCSs based on "reputation" values of individual classifiers: the static reputation-based algorithm (SRB) and the dynamic reputation-based algorithm (DRB). In SRB, each individual classifier is applied to a “validation set”, which is disjoint from training and test sets, to estimate its reputation value. Then, each individual classifier is assigned a weight proportional to its reputation value. Finally, the total decision of the classification system is computed using Bayes rule. We have applied this method to the problem of dysphagia detection in adults with neurogenic swallowing difficulties. The aim was to discriminate between safe and unsafe swallows. The weighted classification accuracy exceeded 85% and, because of its high sensitivity, the SRB approach was deemed suitable for screening purposes. In the next step of this dissertation, I analyzed the SRB algorithm mathematically and examined its asymptotic behavior. Specifically, I contrasted the SRB performance against that of majority voting, the benchmark abstract-level MCS, in the presence of different types of noise. In the second phase of this thesis, I exploited the idea of the Dirichlet reputation system to develop a new MCS method, the dynamic reputation-based algorithm, which is suitable for the classification of non-stationary signals. In this method, the reputation of each classifier is updated dynamically whenever a new sample is classified. At any point in time, a classifier’s reputation reflects the classifier’s performance on both the validation and the test sets. Therefore, the effect of random high-performance of weak classifiers is appropriately moderated and likewise, the effect of a poorly performing individual classifier is mitigated as its reputation value, and hence overall influence on the final decision is diminished. We applied DRB to the challenging problem of discerning physiological responses from nonverbal youth with severe disabilities. The promising experimental results encourage further development of reputation-based multi-classifier systems in the domain of access technology research.

Access technologies often deal with the classification of several physiological and biomechanical signals. In most previous studies involving access technologies, a single classifier has been trained. Despite reported success of these single classifiers, classification accuracies are often below clinically viable levels. One approach to improve upon the performance of these classifiers is to utilize the state of- the-art multiple classifier systems (MCS). Because MCS invoke more than one classifier, more information can be exploited from the signals, potentially leading to higher classification performance than that achievable with single classifiers. Moreover, by decreasing the feature space dimensionality of each classifier, the speed of the system can be increased. MCSs may combine classifiers on three levels: abstract, rank, or measurement level. Among them, abstract-level MCSs have been the most widely applied in the literature given the flexibility of the abstract level output, i.e., class labels may be derived from any type of classifier and outputs from multiple classifiers, each designed within a different context, can be easily combined. In this thesis, we develop two new abstract-level MCSs based on "reputation" values of individual classifiers: the static reputation-based algorithm (SRB) and the dynamic reputation-based algorithm (DRB). In SRB, each individual classifier is applied to a “validation set”, which is disjoint from training and test sets, to estimate its reputation value. Then, each individual classifier is assigned a weight proportional to its reputation value. Finally, the total decision of the classification system is computed using Bayes rule. We have applied this method to the problem of dysphagia detection in adults with neurogenic swallowing difficulties. The aim was to discriminate between safe and unsafe swallows. The weighted classification accuracy exceeded 85% and, because of its high sensitivity, the SRB approach was deemed suitable for screening purposes. In the next step of this dissertation, I analyzed the SRB algorithm mathematically and examined its asymptotic behavior. Specifically, I contrasted the SRB performance against that of majority voting, the benchmark abstract-level MCS, in the presence of different types of noise. In the second phase of this thesis, I exploited the idea of the Dirichlet reputation system to develop a new MCS method, the dynamic reputation-based algorithm, which is suitable for the classification of non-stationary signals. In this method, the reputation of each classifier is updated dynamically whenever a new sample is classified. At any point in time, a classifier’s reputation reflects the classifier’s performance on both the validation and the test sets. Therefore, the effect of random high-performance of weak classifiers is appropriately moderated and likewise, the effect of a poorly performing individual classifier is mitigated as its reputation value, and hence overall influence on the final decision is diminished. We applied DRB to the challenging problem of discerning physiological responses from nonverbal youth with severe disabilities. The promising experimental results encourage further development of reputation-based multi-classifier systems in the domain of access technology research.

<p>In the recent years a growing interest in Cognitive Science has been directed to the cognitive role of the agent's ability to predict the consequences of their actions, without actual engagement with their environment. The creation of an experimental model for Hesslow's simulation hypothesis, based on the use of a simulated adaptive agent and the methods of evolutionary robotics within the general perspective of radical connectionism, is reported in this dissertation. A hierarchical architecture consisting of a mixture of (recurrent) experts is investigated in order to test its ability to produce an 'inner world', functional stand-in for the agent's interactions with its environment. Such a mock world is expected to be rich enough to sustain 'blind navigation', which means navigation based solely on the agent's own internal predictions. The results exhibit the system's vivid internal dynamics, its critical sensitivity to a high number of parameters and, finally, a discrepancy with the declared goal of blind navigation. However, given the dynamical complexity of the system, further analysis and testing appear necessary.</p>

In the recent years a growing interest in Cognitive Science has been directed to the cognitive role of the agent's ability to predict the consequences of their actions, without actual engagement with their environment. The creation of an experimental model for Hesslow's simulation hypothesis, based on the use of a simulated adaptive agent and the methods of evolutionary robotics within the general perspective of radical connectionism, is reported in this dissertation. A hierarchical architecture consisting of a mixture of (recurrent) experts is investigated in order to test its ability to produce an 'inner world', functional stand-in for the agent's interactions with its environment. Such a mock world is expected to be rich enough to sustain 'blind navigation', which means navigation based solely on the agent's own internal predictions. The results exhibit the system's vivid internal dynamics, its critical sensitivity to a high number of parameters and, finally, a discrepancy with the declared goal of blind navigation. However, given the dynamical complexity of the system, further analysis and testing appear necessary.

This thesis proposes novel algorithms for using Gaussian processes for Discriminative pose estimation. We overcome the traditional limitations of Gaussian processes, their cubic training complexity and their uni-modal predictive distribution by assembling them in a mixture of experts formulation. Our First contribution shows that by creating a large number of Fixed size Gaussian process experts, we can build a model that is able to scale to large data sets and accurately learn the multi-modal and non- linear mapping between image features and the subject’s pose. We demonstrate that this model gives state of the art performance compared to other discriminative pose estimation techniques.We then extend the model to automatically learn the size and location of each expert. Gaussian processes are able to accurately model non-linear functional regression problems where the output is given as a function of the input. However, when an individual Gaussian process is trained on data which contains multi-modalities, or varying levels of ambiguity, the Gaussian process is unable to accurately model the data. We propose a novel algorithm for learning the size and location of each expert in our mixture of Gaussian processes model to ensure that the training data of each expert matches the assumptions of a Gaussian process. We show that this model is able to out perform our previous mixture of Gaussian processes model.Our final contribution is a dynamics framework for inferring a smooth sequence of pose estimates from a sequence of independent predictive distributions. Discriminative pose estimation infers the pose of each frame independently, leading to jittery tracking results. Our novel algorithm uses a model of human dynamics to infer a smooth path through a sequence of Gaussian mixture models as given by our mixture of Gaussian processes model. We show that our algorithm is able to smooth and correct some mis- takes made by the appearance model alone, and outperform a baseline linear dynamical system.

No description available.

A method for implicit variable selection in mixture-of-experts frameworks is proposed. We introduce a prior structure where information is taken from a set of independent covariates. Robust class membership predictors are identified using a normal gamma prior. The resulting model setup is used in a finite mixture of Bernoulli distributions to find homogenous clusters of women in Mozambique based on their information sources on HIV. Fully Bayesian inference is carried out via the implementation of a Gibbs sampler.