A Stochastic View of Optimal Regret through Minimax Duality

We study the regret of optimal strategies for online convex optimization games. Using von Neumann's minimax theorem, we show that the optimal regret in this adversarial setting is closely related to the behavior of the empirical minimization algorithm in a stochastic process setting: it is equal to the maximum, over joint distributions of the adversary's action sequence, of the difference between a sum of minimal expected losses and the minimal empirical loss. We show that the optimal regret has a natural geometric interpretation, since it can be viewed as the gap in Jensen's inequality for a concave functional--the minimizer over the player's actions of expected loss--defined on a set of probability distributions. We use this expression to obtain upper and lower bounds on the regret of an optimal strategy for a variety of online learning problems. Our method provides upper bounds without the need to construct a learning algorithm; the lower bounds provide explicit optimal strategies for the adversary. Peter L. Bartlett, Alexander Rakhlin

Making coups history

International support is capable of making the difference between the successful defense of democracy and its ignominious defeat. Indeed, the perceived probability of both support for democratically chosen leaders and opposition to their attackers can fundamentally shift the balance in the domestic struggle between them. Nevertheless, although changes to international law and international relations justify a greater international role in preventing and deterring coups and erosions, not all responsibility for protecting democracy should be assigned to the international community. Indeed, the first line of defense should be a democracy’s own domestic initiatives, with the main role of the international community being to support a domestic response to threats to democracy.

Seeing the Wood for the Trees: A Critical Evaluation of Methods to Estimate the Parameters of Stochastic Differential Equations

Maximum-likelihood estimates of the parameters of stochastic differential equations are consistent and asymptotically efficient, but unfortunately difficult to obtain if a closed-form expression for the transitional probability density function of the process is not available. As a result, a large number of competing estimation procedures have been proposed. This article provides a critical evaluation of the various estimation techniques. Special attention is given to the ease of implementation and comparative performance of the procedures when estimating the parameters of the Cox–Ingersoll–Ross and Ornstein–Uhlenbeck equations respectively.

Hierarchical Bayesian modelling of early detection surveillance for plant pest invasions

Early detection surveillance programs aim to find invasions of exotic plant pests and diseases before they are too widespread to eradicate. However, the value of these programs can be difficult to justify when no positive detections are made. To demonstrate the value of pest absence information provided by these programs, we use a hierarchical Bayesian framework to model estimates of incursion extent with and without surveillance. A model for the latent invasion process provides the baseline against which surveillance data are assessed. Ecological knowledge and pest management criteria are introduced into the model using informative priors for invasion parameters. Observation models assimilate information from spatio-temporal presence/absence data to accommodate imperfect detection and generate posterior estimates of pest extent. When applied to an early detection program operating in Queensland, Australia, the framework demonstrates that this typical surveillance regime provides a modest reduction in the estimate that a surveyed district is infested. More importantly, the model suggests that early detection surveillance programs can provide a dramatic reduction in the putative area of incursion and therefore offer a substantial benefit to incursion management. By mapping spatial estimates of the point probability of infestation, the model identifies where future surveillance resources can be most effectively deployed.

Statistical analyses of weigh-in-motion data for bridge live load development

This paper discusses the statistical analyses used to derive bridge live loads models for Hong Kong from a 10-year weigh-in-motion (WIM) data. The statistical concepts required and the terminologies adopted in the development of bridge live load models are introduced. This paper includes studies for representative vehicles from the large amount of WIM data in Hong Kong. Different load affecting parameters such as gross vehicle weights, axle weights, axle spacings, average daily number of trucks etc are first analyzed by various stochastic processes in order to obtain the mathematical distributions of these parameters. As a prerequisite to determine accurate bridge design loadings in Hong Kong, this study not only takes advantages of code formulation methods used internationally but also presents a new method for modelling collected WIM data using a statistical approach.

Fitness and fatness in childhood obesity : implications of physical activity

Overweight and obesity are a significant cause of poor health worldwide, particularly in conjunction with low levels of physical activity (PA). PA is health-protective and essential for the physical growth and development of children, promoting physical and psychological health while simultaneously increasing the probability of remaining active as an adult. However, many obese children and adolescents have a unique set of physiological, biomechanical, and neuromuscular barriers to PA that they must overcome. It is essential to understand the influence of these barriers on an obese child's motivation in order to exercise and tailor exercise programs to the special needs of this population. Chapter Outline • Introduction • Defining Physical Activity, Exercise, and Physical Fitness • Physical Activity, Physical Fitness, And Motor Competence In Obese Children • Physical Activity and Obesity in Children • Physical Fitness in Obese Children • Balance and Gait in Obese Children • Motor Competence in Obese Children • Physical Activity Guidelines for Obese Children • Clinical Assessment of the Obese Child • Physical Activity Characteristics: Mode • Physical Activity Characteristics: Intensity • Physical Activity Characteristics: Frequency • Physical Activity Characteristics: Duration • Conclusion

Protection of distributed generation interfaced networks

With the rapid increase in electrical energy demand, power generation in the form of distributed generation is becoming more important. However, the connections of distributed generators (DGs) to a distribution network or a microgrid can create several protection issues. The protection of these networks using protective devices based only on current is a challenging task due to the change in fault current levels and fault current direction. The isolation of a faulted segment from such networks will be difficult if converter interfaced DGs are connected as these DGs limit their output currents during the fault. Furthermore, if DG sources are intermittent, the current sensing protective relays are difficult to set since fault current changes with time depending on the availability of DG sources. The system restoration after a fault occurs is also a challenging protection issue in a converter interfaced DG connected distribution network or a microgrid. Usually, all the DGs will be disconnected immediately after a fault in the network. The safety of personnel and equipment of the distribution network, reclosing with DGs and arc extinction are the major reasons for these DG disconnections. In this thesis, an inverse time admittance (ITA) relay is proposed to protect a distribution network or a microgrid which has several converter interfaced DG connections. The ITA relay is capable of detecting faults and isolating a faulted segment from the network, allowing unfaulted segments to operate either in grid connected or islanded mode operations. The relay does not make the tripping decision based on only the fault current. It also uses the voltage at the relay location. Therefore, the ITA relay can be used effectively in a DG connected network in which fault current level is low or fault current level changes with time. Different case studies are considered to evaluate the performance of the ITA relays in comparison to some of the existing protection schemes. The relay performance is evaluated in different types of distribution networks: radial, the IEEE 34 node test feeder and a mesh network. The results are validated through PSCAD simulations and MATLAB calculations. Several experimental tests are carried out to validate the numerical results in a laboratory test feeder by implementing the ITA relay in LabVIEW. Furthermore, a novel control strategy based on fold back current control is proposed for a converter interfaced DG to overcome the problems associated with the system restoration. The control strategy enables the self extinction of arc if the fault is a temporary arc fault. This also helps in self system restoration if DG capacity is sufficient to supply the load. The coordination with reclosers without disconnecting the DGs from the network is discussed. This results in increased reliability in the network by reduction of customer outages.

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.

Empirical minimization

We investigate the behavior of the empirical minimization algorithm using various methods. We first analyze it by comparing the empirical, random, structure and the original one on the class, either in an additive sense, via the uniform law of large numbers, or in a multiplicative sense, using isomorphic coordinate projections. We then show that a direct analysis of the empirical minimization algorithm yields a significantly better bound, and that the estimates we obtain are essentially sharp. The method of proof we use is based on Talagrand’s concentration inequality for empirical processes.

On the optimality of sample-based estimates of the expectation of the empirical minimizer

We study sample-based estimates of the expectation of the function produced by the empirical minimization algorithm. We investigate the extent to which one can estimate the rate of convergence of the empirical minimizer in a data dependent manner. We establish three main results. First, we provide an algorithm that upper bounds the expectation of the empirical minimizer in a completely data-dependent manner. This bound is based on a structural result due to Bartlett and Mendelson, which relates expectations to sample averages. Second, we show that these structural upper bounds can be loose, compared to previous bounds. In particular, we demonstrate a class for which the expectation of the empirical minimizer decreases as O(1/n) for sample size n, although the upper bound based on structural properties is Ω(1). Third, we show that this looseness of the bound is inevitable: we present an example that shows that a sharp bound cannot be universally recovered from empirical data.

Classification with a reject option using a hinge loss

We consider the problem of binary classification where the classifier can, for a particular cost, choose not to classify an observation. Just as in the conventional classification problem, minimization of the sample average of the cost is a difficult optimization problem. As an alternative, we propose the optimization of a certain convex loss function φ, analogous to the hinge loss used in support vector machines (SVMs). Its convexity ensures that the sample average of this surrogate loss can be efficiently minimized. We study its statistical properties. We show that minimizing the expected surrogate loss—the φ-risk—also minimizes the risk. We also study the rate at which the φ-risk approaches its minimum value. We show that fast rates are possible when the conditional probability P(Y=1|X) is unlikely to be close to certain critical values.

Sparseness vs estimating conditional probabilities : some asymptotic results

One of the nice properties of kernel classifiers such as SVMs is that they often produce sparse solutions. However, the decision functions of these classifiers cannot always be used to estimate the conditional probability of the class label. We investigate the relationship between these two properties and show that these are intimately related: sparseness does not occur when the conditional probabilities can be unambiguously estimated. We consider a family of convex loss functions and derive sharp asymptotic results for the fraction of data that becomes support vectors. This enables us to characterize the exact trade-off between sparseness and the ability to estimate conditional probabilities for these loss functions.

Adaboost is consistent

The risk, or probability of error, of the classifier produced by the AdaBoost algorithm is investigated. In particular, we consider the stopping strategy to be used in AdaBoost to achieve universal consistency. We show that provided AdaBoost is stopped after n1-ε iterations---for sample size n and ε ∈ (0,1)---the sequence of risks of the classifiers it produces approaches the Bayes risk.