In many personalized recommendation scenarios, the generalization ability of a target task can be improved via learning with additional auxiliary tasks alongside this target task on a multi-task network. However, this method often suffers from a serious optimization imbalance problem. On the one hand, one or more auxiliary tasks might have a larger influence than the target task and even dominate the network weights, resulting in worse recommendation accuracy for the target task. On the other hand, the influence of one or more auxiliary tasks might be too weak to assist the target task. More challenging is that this imbalance dynamically changes throughout the training process and varies across the parts of the same network. We propose a new method: MetaBalance to balance auxiliary losses via directly manipulating their gradients w.r.t the shared parameters in the multi-task network. Specifically, in each training iteration and adaptively for each part of the network, the gradient of an auxiliary loss is carefully reduced or enlarged to have a closer magnitude to the gradient of the target loss, preventing auxiliary tasks from being so strong that dominate the target task or too weak to help the target task. Moreover, the proximity between the gradient magnitudes can be flexibly adjusted to adapt MetaBalance to different scenarios. The experiments show that our proposed method achieves a significant improvement of 8.34% in terms of [email protected] upon the strongest baseline on two real-world datasets. The code of our approach can be found at here.1

Network alignment plays an important role in a variety of applications. Many traditional methods explicitly or implicitly assume the alignment consistency which might suffer from over-smoothness, whereas some recent embedding based methods could somewhat embrace the alignment disparity by sampling negative alignment pairs. However, under different or even competing designs of negative sampling distributions, some methods advocate positive correlation which could result in false negative samples incorrectly violating the alignment consistency, whereas others champion negative correlation or uniform distribution to sample nodes which may contribute little to learning meaningful embeddings. In this paper, we demystify the intrinsic relationships behind various network alignment methods and between these competing design principles of sampling. Specifically, in terms of model design, we theoretically reveal the close connections between a special graph convolutional network model and the traditional consistency based alignment method. For model training, we quantify the risk of embedding learning for network alignment with respect to the sampling distributions. Based on these, we propose NeXtAlign which strikes a balance between alignment consistency and disparity. We conduct extensive experiments that demonstrate the proposed method achieves significant improvements over the state-of-the-arts.

Understanding content at scale is a difficult but important problem for many platforms. Many previous studies focus on content understanding to optimize engagement with existing users. However, little work studies how to leverage better content understanding to attract new users. In this work, we build a framework for generating natural language content annotations and show how they can be used for search engine optimization. The proposed framework relies on an XGBoost model that labels “pins” with high probability phrases, and a logistic regression layer that learns to rank aggregated annotations for groups of content. The pipeline identifies keywords that are descriptive and contextually meaningful. We perform a large-scale production experiment deployed on the Pinterest platform and show that natural language annotations cause a 1-2% increase in traffic from leading search engines. This increase is statistically significant. Finally, we explore and interpret the characteristics of our annotations framework.

In this paper, we survey the current state-of-art models for structured learning problems, including Hidden Markov Model (HMM), Conditional Random Fields (CRF), Averaged Perceptron (AP), Structured SVMs (SVMstruct), Max Margin Markov Networks (M3N), and an integration of search and learning algorithm (SEARN). With all due tuning efforts of various parameters of each model, on the data sets we have applied the models to, we found that SVMstruct enjoys better performance compared with the others. In addition, we also propose a new method which we call the Structured Learning Ensemble (SLE) to combine these structured learning models. Empirical results show that our SLE algorithm provides more accurate solutions compared with the best results of the individual models.

In this paper we developed an Inductive Logic Programming (ILP) based framework ExOpaque that is able to extract a set of Horn clauses from an arbitrary opaque machine learning model, to describe the behavior of the opaque model with high fidelity while maintaining the simplicity of the Horn clauses for human interpretations.

In this paper, a combinatorial auction problem is modeled as a NP-complete set packing problem and a Lagrangian relaxation based heuristic algorithm is proposed. Extensive experiments are conducted using benchmark CATS test sets and more complex test sets. The algorithm provides optimal solutions for most test sets and is always 1%from the optimal solutions for all CATS test sets. Comparisons with CPLEX 8.0 are also provided, which show that the algorithm provides good solutions

This paper extends carrier assignment models used in bid analysis for transportation procurement to incorporate shipper business considerations. These include restricting carrier numbers, favoring incumbents and performance considerations. We provide representative models and develop solutions for these which include the use of metaheuristics. Experimentation shows that our algorithms work well.