A growing proportion of digital advertising slots is purchased through real time bidding auctions, which enables advertisers to impose highly specific criteria on which devices and opportunities to target. Employing sophisticated targeting criteria reliably increases the performance of an ad campaign, however too strict criteria will limit its scale. This raises the need to estimate the number of anticipated ad impressions at a given campaign performance level, thus enabling advertisers to tune the campaign's budget to an optimal performance-scale trade-off. In this paper, we provide a way to estimate campaign impressions given the campaign criteria. There are several challenges to this problem. First, the criteria contain logic to include and exclude combinations of audience segments, making the space of possible criteria exponentially large. Furthermore, it is difficult to validate predictions, because we wish to predict the number of impressions available without budget constraints, a situation we can rarely observe in practice. In our approach, we first treat the audience segment inclusion/exclusion criteria separately as a data compression problem, where we use MinHash "sketches" to estimate audience size. We then model the number of available impressions with a regularized linear regression in log space, using multiplier features motivated by the assumption that some components of the additional campaign criteria are conditionally independent. We construct a validation set by projecting observed RTB data (under real budget constraints) to get impression availability without budget constraints. Using this approach, our average prediction is a factor of 2.2 from the true impression availability, and the deployed product responds to user requests in well under a second, meeting both accuracy and latency requirements for decision making in the execution of advertising campaigns.

Machine Learning research is progressing at an ever increasing pace. Fueled by technology advances commonly referred to as "Big Data", all data related fields are teaming with scientific and applied activity: our communities explore new application areas, develop new learning algorithms, and continuously scale and improve optimization and estimation methods. But from an industry perspective, many of the most impeding challenges are entirely elsewhere. This talk takes a fresh look at the practical state of affairs in the context of running a large-scale automated machine learning system that supports 50 Billion decision daily on behalf of hundreds of digital advertisers. Some of the key lessons are 1) robustness beats peak performance almost always, 2) support for the constant dynamic fluctuations in the data stream is essential, 3) models exploiting unknowingly any weakness of your metrics, and finally 4) the fact that despite big data, the data you really want never exists.

Internet display advertising is a critical revenue source for publishers and online content providers, and is supported by massive amounts of user and publisher data. Targeting display ads can be improved substantially with machine learning methods, but building many models on massive data becomes prohibitively expensive computationally. This paper presents a combination of strategies, deployed by the online advertising firm Dstillery, for learning many models from extremely high-dimensional data efficiently and without human intervention. This combination includes: (i)~A method for simple-yet-effective transfer learning where a model learned from data that is relatively abundant and cheap is taken as a prior for Bayesian logistic regression trained with stochastic gradient descent (SGD) from the more expensive target data. (ii)~A new update rule for automatic learning rate adaptation, to support learning from sparse, high-dimensional data, as well as the integration with adaptive regularization. We present an experimental analysis across 100 different ad campaigns, showing that the transfer learning indeed improves performance across a large number of them, especially at the start of the campaigns. The combined "hands-free" method needs no fiddling with the SGD learning rate, and we show that it is just as effective as using expensive grid search to set the regularization parameter for each campaign.

Microsoft Academic Search is a free search engine specific to scholarly material. It currently covers more than 50 million publications and over 19 million authors across a variety of domains. One of the main challenges in correctly indexing this material is author name ambiguity and the resulting noise in author profiles. KDD Cup 2013 invited participants to tackle this problem in 2 ways: (1) by automatically determining which papers in an author profile are truly written by a given author, and (2) by identifying which author profiles need to be merged because they belong to the same author. This paper presents a brief account of the contest and the lessons learned.

The automated targeting of online display ads at scale requires the simultaneous evaluation of a single prospect against many independent models. When deciding which ad to show to a user, one must calculate likelihood-to-convert scores for that user across all potential advertisers in the system. For modern machine-learning-based targeting, as conducted by Media6Degrees (M6D), this can mean scoring against thousands of models in a large, sparse feature space. Dimensionality reduction within this space is useful, as it decreases scoring time and model storage requirements. To meet this need, we develop a novel algorithm for scalable supervised dimensionality reduction across hundreds of simultaneous classification tasks. The algorithm performs hierarchical clustering in the space of model parameters from historical models in order to collapse related features into a single dimension. This allows us to implicitly incorporate feature and label data across all tasks without operating directly in a massive space. We present experimental results showing that for this task our algorithm outperforms other popular dimensionality-reduction algorithms across a wide variety of ad campaigns, as well as production results that showcase its performance in practice.

Data generated by observing the actions of web browsers across the internet is being used at an ever increasing rate for both building models and making decisions. In fact, a quarter of the industry-track papers for KDD in 2012 were based on data generated by online actions. The models, analytics and decisions they inform all stem from the assumption that observed data captures the intent of users. However, a large portion of these observed actions are not intentional, and are effectively polluting the models. Much of this observed activity is either generated by robots traversing the internet or the result of unintended actions of real users. These non-intentional actions observed in the web logs severely bias both analytics and the models created from the data. In this paper, we will show examples of how non-intentional traffic that is produced by fraudulent activities adversely affects both general analytics and predictive models, and propose an approach using co-visitation networks to identify sites that have large amounts of this fraudulent traffic. We will then show how this approach, along with a second stage classifier that identifies non-intentional traffic at the browser level, is deployed in production at Media6Degrees (m6d), a targeting technology company for display advertising. This deployed product acts both to filter out the fraudulent traffic from the input data and to insure that we don't serve ads during unintended website visits.

Billions of online display advertising spots are purchased on a daily basis through real time bidding exchanges (RTBs). Advertising companies bid for these spots on behalf of a company or brand in order to purchase these spots to display banner advertisements. These bidding decisions must be made in fractions of a second after the potential purchaser is informed of what location (Internet site) has a spot available and who would see the advertisement. The entire transaction must be completed in near real-time to avoid delays loading the page and maintain a good users experience. This paper presents a bid-optimization approach that is implemented in production at Media6Degrees for bidding on these advertising opportunities at an appropriate price. The approach combines several supervised learning algorithms, as well as second price auction theory, to determine the correct price to ensure that the right message is delivered to the right person, at the right time.

Most data mining research is concerned with building high-quality classification models in isolation. In massive production systems, however, the ability to monitor and maintain performance over time while growing in size and scope is equally important. Many external factors may degrade classification performance including changes in data distribution, noise or bias in the source data, and the evolution of the system itself. A well-functioning system must gracefully handle all of these. This paper lays out a set of design principles for large-scale autonomous data mining systems and then demonstrates our application of these principles within the m6d automated ad targeting system. We demonstrate a comprehensive set of quality control processes that allow us monitor and maintain thousands of distinct classification models automatically, and to add new models, take on new data, and correct poorly-performing models without manual intervention or system disruption.

Deemed “one of the top ten data mining mistakes”, leakage is the introduction of information about the data mining target that should not be legitimately available to mine from. In addition to our own industry experience with real-life projects, controversies around several major public data mining competitions held recently such as the INFORMS 2010 Data Mining Challenge and the IJCNN 2011 Social Network Challenge are evidence that this issue is as relevant today as it has ever been. While acknowledging the importance and prevalence of leakage in both synthetic competitions and real-life data mining projects, existing literature has largely left this idea unexplored. What little has been said turns out not to be broad enough to cover more complex cases of leakage, such as those where the classical independently and identically distributed (i.i.d.) assumption is violated, that have been recently documented. In our new approach, these cases and others are explained by explicitly defining modeling goals and analyzing the broader framework of the data mining problem. The resulting definition enables us to derive general methodology for dealing with the issue. We show that it is possible to avoid leakage with a simple specific approach to data management followed by what we call a learn-predict separation, and present several ways of detecting leakage when the modeler has no control over how the data have been collected. We also offer an alternative point of view on leakage that is based on causal graph modeling concepts.

Deemed "one of the top ten data mining mistakes", leakage is essentially the introduction of information about the data mining target, which should not be legitimately available to mine from. In addition to our own industry experience with real-life projects, controversies around several major public data mining competitions held recently such as the INFORMS 2010 Data Mining Challenge and the IJCNN 2011 Social Network Challenge are evidence that this issue is as relevant today as it has ever been. While acknowledging the importance and prevalence of leakage in both synthetic competitions and real-life data mining projects, existing literature has largely left this idea unexplored. What little has been said turns out not to be broad enough to cover more complex cases of leakage, such as those where the classical i.i.d. assumption is violated, that have been recently documented. In our new approach, these cases and others are explained by explicitly defining modeling goals and analyzing the broader framework of the data mining problem. The resulting definition enables us to derive general methodology for dealing with the issue. We show that it is possible to avoid leakage with a simple specific approach to data management followed by what we call a learn-predict separation, and present several ways of detecting leakage when the modeler has no control over how the data have been collected.

The number of patents filed each year has increased dramatically in recent years, raising concerns that patents of questionable validity are restricting the issuance of truly innovative patents. For this reason, there is a strong demand to develop an objective model to quantify patent quality and characterize the attributes that lead to higher-quality patents. In this paper, we develop a latent graphical model to infer patent quality from related measurements. In addition, we extract advanced lexical features via natural language processing techniques to capture the quality measures such as clarity of claims, originality, and importance of cited prior art. We demonstrate the effectiveness of our approach by validating its predictions with previous court decisions of litigated patents.

Attribution of climate change to causal factors has been based predominantly on simulations using physical climate models, which have inherent limitations in describing such a complex and chaotic system. We propose an alternative, data centric, approach that relies on actual measurements of climate observations and human and natural forcing factors. Specifically, we develop a novel method to infer causality from spatial-temporal data, as well as a procedure to incorporate extreme value modeling into our method in order to address the attribution of extreme climate events, such as heatwaves. Our experimental results on a real world dataset indicate that changes in temperature are not solely accounted for by solar radiance, but attributed more significantly to CO2 and other greenhouse gases. Combined with extreme value modeling, we also show that there has been a significant increase in the intensity of extreme temperatures, and that such changes in extreme temperature are also attributable to greenhouse gases. These preliminary results suggest that our approach can offer a useful alternative to the simulation-based approach to climate modeling and attribution, and provide valuable insights from a fresh perspective.

In this paper we discuss the important practical problem of customer wallet estimation, i.e., estimation of potential spending by customers(rather than their expected spending). For this purpose we utilize quantile modeling, whose goal is to estimate a quantile of the discriminative conditional distribution of the response, rather than the mean, which is the implicit goal of most standard regression approaches. We argue that a notion of wallet can be captured through high quantile modeling (e.g, estimating the 90th percentile), and describe a wallet estimation implementation within IBM's Market Alignment Program (MAP). We also discuss the wide range of domains where high-quantile modeling can be practically important: estimating opportunities in sales and marketing domains, defining 'surprising' patterns for outlier and fraud detection and more. We survey some existing approaches for quantile modeling, and propose adaptations of nearest-neighbor and regression-tree approaches to quantile modeling. We demonstrate the various models' performance in high quantile estimation in several domains, including our motivating problem of estimating the 'realistic' IT wallets of IBM customers.

Mass-produced goods tend to be highly standardized in order to maximize manufacturing efficiencies. Some high-value goods with limited production quantities remain much less standardized and each sale can be configured to meet the specific requirements of the customer. In this work we suggest a novel methodology to reduce the number of options for complex product configurations by identifying meaningful sets of options that exhibit strong empirical dependencies in previous customer orders. Our approach explores different measures from statistics and information theory to capture the degree of interdependence between the choices for any pair of product components. We use hierarchical clustering to identify meaningful sets of components that can be combined to decrease the number of unique product specifications and increase production standardization. The focus of our analysis is on the influence of different similarity measure - including chi-squared statistics and versions of mutual information - on the ability of the clustering to find meaningful clusters.

Motivated by the problem of customer wallet estimation, we propose a new setting for multi-view regression, where we learn a completely unobserved target (in our case, customer wallet) by modeling it as a "central link" in a directed graphical model, connecting multiple sets of observed variables. The resulting conditional independence allows us to reduce the maximum discriminative likelihood estimation problem to a convex optimization problem for exponential linear models. We show that under certain modeling assumptions, in particular, when there exist two conditionally independent views and the noise is Gaussian, this problem can be reduced to a single least squares regression. Thus, for this specific, but widely applicable setting, the "unsupervised" multi-view problem can be solved via a simple supervised learning approach. This reduction also allows us to test the statistical independence assumptions underlying the graphical model and perform variable selection. We demonstrate the effectiveness of our approach on our motivating problem of customer wallet estimation and on simulation data.

We suggest the use of ranking-based evaluation measures for regression models, as a complement to the commonly used residual-based evaluation. We argue that in some cases, such as the case study we present, ranking can be the main underlying goal in building a regression model, and ranking performance is the correct evaluation metric. However, even when ranking is not the contextually correct performance metric, the measures we explore still have significant advantages: They are robust against extreme outliers in the evaluation set; and they are interpretable. The two measures we consider correspond closely to non-parametric correlation coefficients commonly used in data analysis (Spearman's p and Kendall's r); and they both have interesting graphical representations, which, similarly to ROC curves, offer useful "partial" model performance views, in addition to a one-number summary in the area under the curve. We illustrate our methods on a case study of evaluating IT wallet size estimation models for IBM's customers.

Model induction from relational data requires aggregation of the values of attributes of related entities. This paper makes three contributions to the study of relational learning. (1) It presents a hierarchy of relational concepts of increasing complexity, using relational schema characteristics such as cardinality, and derives classes of aggregation operators that are needed to learn these concepts. (2) Expanding one level of the hierarchy, it introduces new aggregation operators that model the distributions of the values to be aggregated and (for classification problems) the differences in these distributions by class. (3) It demonstrates empirically on a noisy business domain that more-complex aggregation methods can increase generalization performance. Constructing features using target-dependent aggregations can transform relational prediction tasks so that well-understood feature-vector-based modeling algorithms can be applied successfully.