Controlling chaos? The value and the challenges of applying complexity theory to project management

The literature abounds with descriptions of failures in high-profile projects and a range of initiatives has been generated to enhance project management practice (e.g., Morris, 2006). Estimating from our own research, there are scores of other project failures that are unrecorded. Many of these failures can be explained using existing project management theory; poor risk management, inaccurate estimating, cultures of optimism dominating decision making, stakeholder mismanagement, inadequate timeframes, and so on. Nevertheless, in spite of extensive discussion and analysis of failures and attention to the presumed causes of failure, projects continue to fail in unexpected ways. In the 1990s, three U.S. state departments of motor vehicles (DMV) cancelled major projects due to time and cost overruns and inability to meet project goals (IT-Cortex, 2010). The California DMV failed to revitalize their drivers’ license and registration application process after spending $45 million. The Oregon DMV cancelled their five year, $50 million project to automate their manual, paper-based operation after three years when the estimates grew to $123 million; its duration stretched to eight years or more and the prototype was a complete failure. In 1997, the Washington state DMV cancelled their license application mitigation project because it would have been too big and obsolete by the time it was estimated to be finished. There are countless similar examples of projects that have been abandoned or that have not delivered the requirements.

The sample complexity of pattern classification with neural networks: the size of the weights is more important than the size of the network

Sample complexity results from computational learning theory, when applied to neural network learning for pattern classification problems, suggest that for good generalization performance the number of training examples should grow at least linearly with the number of adjustable parameters in the network. Results in this paper show that if a large neural network is used for a pattern classification problem and the learning algorithm finds a network with small weights that has small squared error on the training patterns, then the generalization performance depends on the size of the weights rather than the number of weights. For example, consider a two-layer feedforward network of sigmoid units, in which the sum of the magnitudes of the weights associated with each unit is bounded by A and the input dimension is n. We show that the misclassification probability is no more than a certain error estimate (that is related to squared error on the training set) plus A3 √((log n)/m) (ignoring log A and log m factors), where m is the number of training patterns. This may explain the generalization performance of neural networks, particularly when the number of training examples is considerably smaller than the number of weights. It also supports heuristics (such as weight decay and early stopping) that attempt to keep the weights small during training. The proof techniques appear to be useful for the analysis of other pattern classifiers: when the input domain is a totally bounded metric space, we use the same approach to give upper bounds on misclassification probability for classifiers with decision boundaries that are far from the training examples.

Fast rates for estimation error and oracle inequalities for model section

We consider complexity penalization methods for model selection. These methods aim to choose a model to optimally trade off estimation and approximation errors by minimizing the sum of an empirical risk term and a complexity penalty. It is well known that if we use a bound on the maximal deviation between empirical and true risks as a complexity penalty, then the risk of our choice is no more than the approximation error plus twice the complexity penalty. There are many cases, however, where complexity penalties like this give loose upper bounds on the estimation error. In particular, if we choose a function from a suitably simple convex function class with a strictly convex loss function, then the estimation error (the difference between the risk of the empirical risk minimizer and the minimal risk in the class) approaches zero at a faster rate than the maximal deviation between empirical and true risks. In this paper, we address the question of whether it is possible to design a complexity penalized model selection method for these situations. We show that, provided the sequence of models is ordered by inclusion, in these cases we can use tight upper bounds on estimation error as a complexity penalty. Surprisingly, this is the case even in situations when the difference between the empirical risk and true risk (and indeed the error of any estimate of the approximation error) decreases much more slowly than the complexity penalty. We give an oracle inequality showing that the resulting model selection method chooses a function with risk no more than the approximation error plus a constant times the complexity penalty.

Overcoming the complexity of diagnostic problems due to sensor network architecture

In fault detection and diagnostics, limitations coming from the sensor network architecture are one of the main challenges in evaluating a system’s health status. Usually the design of the sensor network architecture is not solely based on diagnostic purposes, other factors like controls, financial constraints, and practical limitations are also involved. As a result, it quite common to have one sensor (or one set of sensors) monitoring the behaviour of two or more components. This can significantly extend the complexity of diagnostic problems. In this paper a systematic approach is presented to deal with such complexities. It is shown how the problem can be formulated as a Bayesian network based diagnostic mechanism with latent variables. The developed approach is also applied to the problem of fault diagnosis in HVAC systems, an application area with considerable modeling and measurement constraints.

On the data consumption benefits of accepting increased uncertainty

In the context of learning paradigms of identification in the limit, we address the question: why is uncertainty sometimes desirable? We use mind change bounds on the output hypotheses as a measure of uncertainty and interpret ‘desirable’ as reduction in data memorization, also defined in terms of mind change bounds. The resulting model is closely related to iterative learning with bounded mind change complexity, but the dual use of mind change bounds — for hypotheses and for data — is a key distinctive feature of our approach. We show that situations exist where the more mind changes the learner is willing to accept, the less the amount of data it needs to remember in order to converge to the correct hypothesis. We also investigate relationships between our model and learning from good examples, set-driven, monotonic and strong-monotonic learners, as well as class-comprising versus class-preserving learnability.

The rademacher complexity of coregularized kernel classes

In the multi-view approach to semisupervised learning, we choose one predictor from each of multiple hypothesis classes, and we co-regularize our choices by penalizing disagreement among the predictors on the unlabeled data. We examine the co-regularization method used in the co-regularized least squares (CoRLS) algorithm, in which the views are reproducing kernel Hilbert spaces (RKHS's), and the disagreement penalty is the average squared difference in predictions. The final predictor is the pointwise average of the predictors from each view. We call the set of predictors that can result from this procedure the co-regularized hypothesis class. Our main result is a tight bound on the Rademacher complexity of the co-regularized hypothesis class in terms of the kernel matrices of each RKHS. We find that the co-regularization reduces the Rademacher complexity by an amount that depends on the distance between the two views, as measured by a data dependent metric. We then use standard techniques to bound the gap between training error and test error for the CoRLS algorithm. Experimentally, we find that the amount of reduction in complexity introduced by co regularization correlates with the amount of improvement that co-regularization gives in the CoRLS algorithm.

Narratives of complexity : an ethnographic exploration of resilience and wellbeing among single refugee women in Brisbane

This study aimed to explore resilience and wellbeing among a group of eight refugee women originating from several countries (mainly African) and living in Brisbane, most of whom were single mothers. To challenge mostly quantitative and gender-blind explorations of mental health concepts among refugee groups, the project sought an emic and contextual understanding of resilience and wellbeing. Established perspectives, while useful, tend to overlook the complexities of refugee mental health experiences and can neglect the dense nature of individual stories. The purpose of my study was to contest relatively simplistic narratives of mental health constructs that tend to dominate migrant and refugee studies and influence practice paradigms in the human services field. In this ethnographic exploration of mental health constructs conducted in 2008 and 2009, the use of in-depth interviews, participant observations, and visual ethnographic elements provided an opportunity for refugee women to tell their own stories. The participants’ unique narratives of pre- and post-migration experiences, shaped by specific gender, age, social, cultural and political aspects prevailing in their lives, yielded ‘thick’ ethnographic description (Geertz, 1973) of their social worlds. The findings explored in this study, namely language issues, the impact of community dynamics, and the single status of refugee women, clearly demonstrate that mental health constructs are fluid, multifaceted and complex in reality. In fact, language, community dynamics, and being a single mother, represented both opportunities and barriers in the lives of participants. In some contexts, these factors were conducive to resilience and wellbeing, while in other circumstances, these three elements acted as a hindrance to positive mental health outcomes. There are multiple dimensions to the findings, signifying that the social worlds of refugee women cannot be simplified using set definitions and neat notions of resilience and wellbeing. Instead, the intricacies and complexities embedded in the mundane of the everyday highlight novel conceptualisations of resilience and wellbeing. Based on the particular circumstances of single refugee mothers, whose experiences differ from that of married women, this thesis presents novel articulations of mental health constructs, as an alternative view to existing trends in the literature on refugee issues. Rich and multi-dimensional meanings associated with the socio-cultural determinants of mental health emerged in the process. This thesis’ findings highlight a significant gap in diasporic studies as well as simplistic assumptions about refugee women’s resettlement experiences. Single refugee women’s distinct issues are so complex and dense, that a contextual approach is critical to yield accurate depictions of their circumstances. It is therefore essential to understand refugee lived experiences within broader socio-political contexts to truly appreciate the depth of these narratives. In this manner, critical aspects salient to refugee journeys can inform different understandings of resilience, wellbeing and mental health, and shape contemporary policy and human service practice paradigms.

On a Syntactic Characterization of Classification with a Mind Change Bound

Most learning paradigms impose a particular syntax on the class of concepts to be learned; the chosen syntax can dramatically affect whether the class is learnable or not. For classification paradigms, where the task is to determine whether the underlying world does or does not have a particular property, how that property is represented has no implication on the power of a classifier that just outputs 1’s or 0’s. But is it possible to give a canonical syntactic representation of the class of concepts that are classifiable according to the particular criteria of a given paradigm? We provide a positive answer to this question for classification in the limit paradigms in a logical setting, with ordinal mind change bounds as a measure of complexity. The syntactic characterization that emerges enables to derive that if a possibly noncomputable classifier can perform the task assigned to it by the paradigm, then a computable classifier can also perform the same task. The syntactic characterization is strongly related to the difference hierarchy over the class of open sets of some topological space; this space is naturally defined from the class of possible worlds and possible data of the learning paradigm.

Exploring the process of implementing a smoking cessation intervention in pregnancy training program for midwives

The question of how to implement evidence effectively reveals a deficiency in our knowledge and understanding of the compound factors involved in such a process (Kitson, Rycroft-Malone et al. 2008). Although there is some awareness of the complexities of the process, there has been little exploration of the effectiveness of implementing evidence-based programs in health care. Despite public awareness of the dangers of smoking in pregnancy, and widespread public health measures to prevent smoking-related disease, women still continue to smoke in pregnancy (Ananth, Savitz et al. 1997; Laws and Hilder 2008). Evaluation of public health measures concludes that smoking cessation interventions during pregnancy increase quit rates among pregnant women (Melvin, Dolan-Mullen et al. 2000; Albrecht, Maloni et al. 2004; Lumley, Oliver et al. 2007). Notwithstanding the potential for improvement in health outcomes for pregnant women and their unborn babies, smoking interventions are often conducted poorly or not at all. Although midwives understand why women smoke in pregnancy and parenthood and are aware of the risks of smoking to both the pregnancy and the unborn child, they require specific knowledge and skills in the provision of support and advice on smoking for pregnant women (Bull and Whitehead 2006) . Organisational-change research demonstrates the complexity of the process of planned change in professionalised institutions such as health care (Greenhalgh, Robert et al. 2005). Some innovations and interventions are never accepted, and others are poorly supported (Greenhalgh, Robert et al. 2004). Comprehension of the change process around health promotion is crucial to the implementation of new health promotion interventions within health care (Riley, Taylor et al. 2003). This study utilised a case study approach to explore the process of implementing a smoking cessation training program for midwives in Queensland metropolitan and regional clinical areas, who attended a ‘Train-the-Trainer program’. The study draws on the organisational change work of Greenhalgh et al (2004) as the theoretical framework through which situational and structural factors are explored and examined as they inform the implementation of smoking cessation programs. The research data constituted staged interviews with midwives who instituted training programs for midwives, as well as organisational and policy documentation. Analysis of the data identified some areas that were not fully addressed in the theoretical model; these formed the basis of the Discussion and Implications for Future Research.

A voting mechanism for named entity translation in English-Chinese question answering

In this paper, we describe a voting mechanism for accurate named entity (NE) translation in English–Chinese question answering (QA). This mechanism involves translations from three different sources: machine translation,online encyclopaedia, and web documents. The translation with the highest number of votes is selected. We evaluated this approach using test collection, topics and assessment results from the NTCIR-8 evaluation forum. This mechanism achieved 95% accuracy in NEs translation and 0.3756 MAP in English–Chinese cross-lingual information retrieval of QA.