Tree boosting is a highly effective and widely used machine learning method. In this paper, we describe a scalable end-to-end tree boosting system called XGBoost, which is used widely by data scientists to achieve state-of-the-art results on many machine learning challenges. We propose a novel sparsity-aware algorithm for sparse data and weighted quantile sketch for approximate tree learning. More importantly, we provide insights on cache access patterns, data compression and sharding to build a scalable tree boosting system. By combining these insights, XGBoost scales beyond billions of examples using far fewer resources than existing systems.

/pdf/xgboost-a-scalable-tree-boosting-system-48ocu6x4c7.pdf

XGBoost: A Scalable Tree Boosting System

SciPy is an open source scientific computing library for the Python programming language. SciPy 1.0 was released in late 2017, about 16 years after the original version 0.1 release. SciPy has become a de facto standard for leveraging scientific algorithms in the Python programming language, with more than 600 unique code contributors, thousands of dependent packages, over 100,000 dependent repositories, and millions of downloads per year. This includes usage of SciPy in almost half of all machine learning projects on GitHub, and usage by high profile projects including LIGO gravitational wave analysis and creation of the first-ever image of a black hole (M87). The library includes functionality spanning clustering, Fourier transforms, integration, interpolation, file I/O, linear algebra, image processing, orthogonal distance regression, minimization algorithms, signal processing, sparse matrix handling, computational geometry, and statistics. In this work, we provide an overview of the capabilities and development practices of the SciPy library and highlight some recent technical developments.

/pdf/scipy-1-0-fundamental-algorithms-for-scientific-computing-in-111lr790ol.pdf

SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python

Despite widespread adoption, machine learning models remain mostly black boxes. Understanding the reasons behind predictions is, however, quite important in assessing trust, which is fundamental if one plans to take action based on a prediction, or when choosing whether to deploy a new model. Such understanding also provides insights into the model, which can be used to transform an untrustworthy model or prediction into a trustworthy one. In this work, we propose LIME, a novel explanation technique that explains the predictions of any classifier in an interpretable and faithful manner, by learning an interpretable model locally varound the prediction. We also propose a method to explain models by presenting representative individual predictions and their explanations in a non-redundant way, framing the task as a submodular optimization problem. We demonstrate the flexibility of these methods by explaining different models for text (e.g. random forests) and image classification (e.g. neural networks). We show the utility of explanations via novel experiments, both simulated and with human subjects, on various scenarios that require trust: deciding if one should trust a prediction, choosing between models, improving an untrustworthy classifier, and identifying why a classifier should not be trusted.

/pdf/why-should-i-trust-you-explaining-the-predictions-of-any-ejtlrtykxr.pdf

"Why Should I Trust You?": Explaining the Predictions of Any Classifier

Low-dimensional embeddings of nodes in large graphs have proved extremely useful in a variety of prediction tasks, from content recommendation to identifying protein functions. However, most existing approaches require that all nodes in the graph are present during training of the embeddings; these previous approaches are inherently transductive and do not naturally generalize to unseen nodes. Here we present GraphSAGE, a general, inductive framework that leverages node feature information (e.g., text attributes) to efficiently generate node embeddings for previously unseen data. Instead of training individual embeddings for each node, we learn a function that generates embeddings by sampling and aggregating features from a node's local neighborhood. Our algorithm outperforms strong baselines on three inductive node-classification benchmarks: we classify the category of unseen nodes in evolving information graphs based on citation and Reddit post data, and we show that our algorithm generalizes to completely unseen graphs using a multi-graph dataset of protein-protein interactions.

Inductive Representation Learning on Large Graphs

Knowledge about user goals is crucial for realizing the vision of intelligent agents acting upon user intent on the web. In a departure from existing approaches, this paper proposes a novel approach to the problem of user goal acquisition: The utilization of search query logs for this task. The paper makes the following contributions: (a) it presents an automatic method for the acquisition of user goals from search query logs with useful precision/recall scores (b) it provides insights into the nature and some characteristics of these goals and (c) it shows that the goals acquired from query logs exhibit traits of a long tail distribution.

Acquiring Explicit User Goals from Search Query Logs.

On the web, search engines represent a primary instrument through which users exercise their intent. Understanding the specific goals users express in search queries could improve our theoretical knowledge about strategies for search goal formulation and search behavior, and could equip search engine providers with better descriptions of users' information needs. However, the degree to which goals are explicitly expressed in search queries can be suspected to exhibit considerable variety, which poses a series of challenges for researchers and search engine providers. This paper introduces a novel perspective on analyzing user goals in search query logs by proposing to study different degrees of intentional explicitness. To explore the implications of this perspective, we studied two different degrees of explicitness of user goals in the AOL search query log containing more than 20 million queries. Our results suggest that different degrees of intentional explicitness represent an orthogonal dimension to existing search query categories and that understanding these different degrees is essential for effective search. The overall contribution of this paper is the elaboration of a set of theoretical arguments and empirical evidence that makes a strong case for further studies of different degrees of intentional explicitness in search query logs.

/pdf/different-degrees-of-explicitness-in-intentional-artifacts-869mp8ej7l.pdf

Different degrees of explicitness in intentional artifacts: studying user goals in a large search query log

scikit-learn/scikit-learn: scikit-learn 0.23.2

Cross-lingual adaptation, a special case of domain adaptation, refers to the transfer of classification knowledge between two languages. In this article we describe an extension of Structural Correspondence Learning (SCL), a recently proposed algorithm for domain adaptation, for cross-lingual adaptation. The proposed method uses unlabeled documents from both languages, along with a word translation oracle, to induce cross-lingual feature correspondences. From these correspondences a cross-lingual representation is created that enables the transfer of classification knowledge from the source to the target language. The main advantages of this approach over other approaches are its resource efficiency and task specificity. 
We conduct experiments in the area of cross-language topic and sentiment classification involving English as source language and German, French, and Japanese as target languages. The results show a significant improvement of the proposed method over a machine translation baseline, reducing the relative error due to cross-lingual adaptation by an average of 30% (topic classification) and 59% (sentiment classification). We further report on empirical analyses that reveal insights into the use of unlabeled data, the sensitivity with respect to important hyperparameters, and the nature of the induced cross-lingual correspondences.

Peter Prettenhofer

Papers

Acquiring Explicit User Goals from Search Query Logs.

Different degrees of explicitness in intentional artifacts: studying user goals in a large search query log

scikit-learn/scikit-learn: scikit-learn 0.23.2

Cross-Lingual Adaptation using Structural Correspondence Learning

scikit-learn/scikit-learn: Scikit-learn 0.22.1