Relevance and diversity are both crucial criteria for an effective search system. In this paper, we propose a unified learning framework for simultaneously optimizing both relevance and diversity. Specifically, the problem is formalized as a structural learning framework optimizing DiversityCorrelated Evaluation Measures (DCEM), such as ERR-IA, α-NDCG and NRBP. Within this framework, the discri...

This paper introduces a novel wrapper-based outlier detection framework (WrapperOD) and its instance (HOUR) for identifying outliers in noisy data (i.e., data with noisy features) with strong couplings between outlying behaviors. Existing subspace or feature selection-based methods are significantly challenged by such data, as their search of feature subset(s) is independent of outlier scoring ...

We show several high-probability concentration bounds for learning unigrams language model. One interesting quantity is the probability of all words appearing exactly k times in a sample of size m. A standard estimator for this quantity is the Good-Turing estimator. The existing analysis on its error shows a high-probability bound of approximately O ( k √ m ) . We improve its dependency on k to...

This manuscript shows that AdaBoost and its immediate variants can produce approximate maximum margin classifiers simply by scaling step size choices with a fixed small constant. In this way, when the unscaled step size is an optimal choice, these results provide guarantees for Friedman’s empirically successful “shrinkage” procedure for gradient boosting (Friedman, 2000). Guarantees are also pr...

A successful approach to structured learning is to write the learning objective as a joint function of linear parameters and inference messages, and iterate between updates to each. This paper observes that if the inference problem is “smoothed” through the addition of entropy terms, for fixed messages, the learning objective reduces to a traditional (non-structured) logistic regression problem...

We present simulation results for supercontinuum generation using tapered As2S3 chalcogenide photonic crystal fibers (PCFs). We demonstrate that an increased soliton self-frequency shift can be achieved using a tapered PCF. There is an optimal tapered PCF, which yields an additional 0:4 mm shift to longer wavelengths relative to the shift that is obtained in an untapered PCF, leading to an incr...

Event Detection Identifying patterns of interest in large temporal datasets Spatial Scan Statistic A method for identifying hotspots in spatial data, widely used in epidemiology and biosurveillance Scoring Function An objective function that measures the anomalousness of a subset of data LTSS Linear-time subset scanning Time to Detect Evaluation metric; time delay before detecting an event Over...

We model how decisions about the allocation of available funds to loans and Treasuries are dependent on perceptions about risk and regret in the banking industry. Our discussion is based on utility theory where a regret attribute is considered alongside a risk component as an integral part of the objective function. In addition, we provide a comparison between a riskand a regret-averse banks. I...

We consider prediction with expert advice under the log-loss with the goal of deriving efficient and robust algorithms. We argue that existing algorithms such as exponentiated gradient, online gradient descent and online Newton step do not adequately satisfy both requirements. Our main contribution is an analysis of the Prod algorithm that is robust to any data sequence and runs in linear time ...

Aggregate loss functions are loss functions for counts of areas where the count itself is of interest. Proportional loss functions are loss functions for counts of areas where the count as a proportion of a total is of interest. To date, much of the work on adjustment has concentrated on aggregate loss functions. A question is whether this effort will necessarily produce estimated counts whi.ch...