Improving LVCSR system combination using neural network language model cross adaptation

State-of-the-art large vocabulary continuous speech recognition (LVCSR) systems often combine outputs from multiple subsystems developed at different sites. Cross system adaptation can be used as an alternative to direct hypothesis level combination schemes such as ROVER. The standard approach involves only cross adapting acoustic models. To fully exploit the complimentary features among sub-systems, language model (LM) cross adaptation techniques can be used. Previous research on multi-level n-gram LM cross adaptation is extended to further include the cross adaptation of neural network LMs in this paper. Using this improved LM cross adaptation framework, significant error rate gains of 4.0%-7.1% relative were obtained over acoustic model only cross adaptation when combining a range of Chinese LVCSR sub-systems used in the 2010 and 2011 DARPA GALE evaluations. Copyright © 2011 ISCA.

Making a shallow network deep: Conversion of a boosting classifier into a decision tree by boolean optimisation

This paper presents a novel way to speed up the evaluation time of a boosting classifier. We make a shallow (flat) network deep (hierarchical) by growing a tree from decision regions of a given boosting classifier. The tree provides many short paths for speeding up while preserving the reasonably smooth decision regions of the boosting classifier for good generalisation. For converting a boosting classifier into a decision tree, we formulate a Boolean optimization problem, which has been previously studied for circuit design but limited to a small number of binary variables. In this work, a novel optimisation method is proposed for, firstly, several tens of variables i.e. weak-learners of a boosting classifier, and then any larger number of weak-learners by using a two-stage cascade. Experiments on the synthetic and face image data sets show that the obtained tree achieves a significant speed up both over a standard boosting classifier and the Fast-exit-a previously described method for speeding-up boosting classification, at the same accuracy. The proposed method as a general meta-algorithm is also useful for a boosting cascade, where it speeds up individual stage classifiers by different gains. The proposed method is further demonstrated for fast-moving object tracking and segmentation problems. © 2011 Springer Science+Business Media, LLC.

State estimation over a communication network: Measurement or estimate communication?

In this paper we consider the problem of state estimation over a communication network. Using estimation quality as a metric, two communication schemes are studied and compared. In scheme one, each sensor node communicates its measurement data to the remote estimator, while in scheme two, each sensor node communicates its local state estimate to the remote estimator. We show that with perfect communication link, if the sensor has unlimited computation capability, the two schemes produce the same estimate at the estimator, and if the sensor has limited computation capability, scheme one is always better than scheme two. On the other hand, when data packet drops occur over the communication link, we show that if the sensor has unlimited computation capability, scheme two always outperforms scheme one, and if the sensor has limited computation capability, we show that in general there exists a critical packet arrival rate, above which scheme one outperforms scheme two. Simulations are provided to demonstrate the two schemes under various circumstances. © South China University of Technology and Academy of Mathematics and Systems Science, CAS and Springer-Verlag Berlin Heidelberg 2010.

Impact of heterogeneous link qualities and network connectivity on binary consensus

In this paper we consider a network that is trying to reach consensus over the occurrence of an event while communicating over Additive White Gaussian Noise (AWGN) channels. We characterize the impact of different link qualities and network connectivity on consensus performance by analyzing both the asymptotic and transient behaviors. More specifically, we derive a tight approximation for the second largest eigenvalue of the probability transition matrix. We furthermore characterize the dynamics of each individual node. © 2009 AACC.

Oscillations in a decentralized economic network

The study of exchange markets dates back to LeonWalras's general equilibrium theory. Since then the economic market has been studied for its' equilibrium properties, fairness of allocations of private and public goods, and even the psychological incentives of participants. This paper studies the dynamics of an exchange economy built on a network of markets where consumers trade with suppliers to optimize utility. Viewing the market in as a decentralized network we study the system from the usual control theory point of view, evaluating the system's dynamic performance, stability and robustness. It is shown that certain consumer demand dynamics can lead to oscillations while others can converge to optimal allocations. © 2011 IFAC.

Robust network reconstruction in polynomial time

This paper presents an efficient algorithm for robust network reconstruction of Linear Time-Invariant (LTI) systems in the presence of noise, estimation errors and unmodelled nonlinearities. The method here builds on previous work [1] on robust reconstruction to provide a practical implementation with polynomial computational complexity. Following the same experimental protocol, the algorithm obtains a set of structurally-related candidate solutions spanning every level of sparsity. We prove the existence of a magnitude bound on the noise, which if satisfied, guarantees that one of these structures is the correct solution. A problem-specific model-selection procedure then selects a single solution from this set and provides a measure of confidence in that solution. Extensive simulations quantify the expected performance for different levels of noise and show that significantly more noise can be tolerated in comparison to the original method. © 2012 IEEE.

Defining product-service network configurations and location roles: A current and future state analysis framework for international engineering operations

Many manufacturing firms have developed a service dimension to their product portfolio. In response to this growing trend of servitisation, organisations, often involved in complex, long-lifecycle product-service system (PSS) provision, need to reconfigure their global engineering networks to support integrated PSS offerings. Drawing on parallel concepts in 'production' networks, the idea of 'location role' now becomes increasingly complex, in terms of service delivery. As new markets develop, locations in a specific region may need to grow/adapt engineering service 'competencies' along the value chain, from design and build to support and service, in order to serve future location-specific requirements and, potentially, those requirements of the overall network. The purpose of this paper is to advance understanding of how best to design complex multi-organisational engineering service networks, through extension of the 'production' network location role concept to a PSS context, capturing both traditional engineering 'design and build' and engineering 'service' requirements. Copyright © 2012 Inderscience Enterprises Ltd.

A dependent partition-valued process for multitask clustering and time evolving network modelling

The fundamental aim of clustering algorithms is to partition data points. We consider tasks where the discovered partition is allowed to vary with some covariate such as space or time. One approach would be to use fragmentation-coagulation processes, but these, being Markov processes, are restricted to linear or tree structured covariate spaces. We define a partition-valued process on an arbitrary covariate space using Gaussian processes. We use the process to construct a multitask clustering model which partitions datapoints in a similar way across multiple data sources, and a time series model of network data which allows cluster assignments to vary over time. We describe sampling algorithms for inference and apply our method to defining cancer subtypes based on different types of cellular characteristics, finding regulatory modules from gene expression data from multiple human populations, and discovering time varying community structure in a social network.