RaSE: A Variable Screening Framework via Random Subspace Ensembles

Variable screening methods have been shown to be effective in dimension reduction under the ultra-high dimensional setting. Most existing screening methods are designed to rank the predictors according to their individual contributions to the response. As a result, variables that are marginally independent but jointly dependent with the response could be missed. In this work, we propose a new framework for variable screening, Random Subspace Ensemble (RaSE), which works by evaluating the quality of random subspaces that may cover multiple predictors. This new screening framework can be naturally combined with any subspace evaluation criterion, which leads to an array of screening methods. The framework is capable to identify signals with no marginal effect or with high-order interaction effects. It is shown to enjoy the sure screening property and rank consistency. We also develop an iterative version of RaSE screening with theoretical support. Extensive simulation studies and real-data analysis show the effectiveness of the new screening framework.

WeChat AI's Submission for DSTC9 Interactive Dialogue Evaluation Track

We participate in the DSTC9 Interactive Dialogue Evaluation Track (Gunasekara et al. 2020) sub-task 1 (Knowledge Grounded Dialogue) and sub-task 2 (Interactive Dialogue). In sub-task 1, we employ a pre-trained language model to generate topic-related responses and propose a response ensemble method for response selection. In sub-task2, we propose a novel Dialogue Planning Model (DPM) to capture conversation flow in the interaction with humans. We also design an integrated open-domain dialogue system containing pre-process, dialogue model, scoring model, and post-process, which can generate fluent, coherent, consistent, and humanlike responses. We tie 1st on human ratings and also get the highest Meteor, and Bert-score in sub-task 1, and rank 3rd on interactive human evaluation in sub-task 2.

The Interplay of Demographic Variables and Social Distancing Scores in Deep Prediction of U.S. COVID-19 Cases

With the severity of the COVID-19 outbreak, we characterize the nature of the growth trajectories of counties in the United States using a novel combination of spectral clustering and the correlation matrix. As the U.S. and the rest of the world are experiencing a severe second wave of infections, the importance of assigning growth membership to counties and understanding the determinants of the growth are increasingly evident. Subsequently, we select the demographic features that are most statistically significant in distinguishing the communities. Lastly, we effectively predict the future growth of a given county with an LSTM using three social distancing scores. This comprehensive study captures the nature of counties' growth in cases at a very micro-level using growth communities, demographic factors, and social distancing performance to help government agencies utilize known information to make appropriate decisions regarding which potential counties to target resources and funding to.

Future-Guided Incremental Transformer for Simultaneous Translation

Simultaneous translation (ST) starts translations synchronously while reading source sentences, and is used in many online scenarios. The previous wait-k policy is concise and achieved good results in ST. However, wait-k policy faces two weaknesses: low training speed caused by the recalculation of hidden states and lack of future source information to guide training. For the low training speed, we propose an incremental Transformer with an average embedding layer (AEL) to accelerate the speed of calculation of the hidden states during training. For future-guided training, we propose a conventional Transformer as the teacher of the incremental Transformer, and try to invisibly embed some future information in the model through knowledge distillation. We conducted experiments on Chinese-English and German-English simultaneous translation tasks and compared with the wait-k policy to evaluate the proposed method. Our method can effectively increase the training speed by about 28 times on average at different k and implicitly embed some predictive abilities in the model, achieving better translation quality than wait-k baseline.

Spectral clustering via adaptive layer aggregation for multi-layer networks

One of the fundamental problems in network analysis is detecting community structure in multi-layer networks, of which each layer represents one type of edge information among the nodes. We propose integrative spectral clustering approaches based on effective convex layer aggregations. Our aggregation methods are strongly motivated by a delicate asymptotic analysis of the spectral embedding of weighted adjacency matrices and the downstream $k$-means clustering, in a challenging regime where community detection consistency is impossible. In fact, the methods are shown to estimate the optimal convex aggregation, which minimizes the mis-clustering error under some specialized multi-layer network models. Our analysis further suggests that clustering using Gaussian mixture models is generally superior to the commonly used $k$-means in spectral clustering. Extensive numerical studies demonstrate that our adaptive aggregation techniques, together with Gaussian mixture model clustering, make the new spectral clustering remarkably competitive compared to several popularly used methods.

Investigating Catastrophic Forgetting During Continual Training for Neural Machine Translation

Neural machine translation (NMT) models usually suffer from catastrophic forgetting during continual training where the models tend to gradually forget previously learned knowledge and swing to fit the newly added data which may have a different distribution, e.g. a different domain. Although many methods have been proposed to solve this problem, we cannot get to know what causes this phenomenon yet. Under the background of domain adaptation, we investigate the cause of catastrophic forgetting from the perspectives of modules and parameters (neurons). The investigation on the modules of the NMT model shows that some modules have tight relation with the general-domain knowledge while some other modules are more essential in the domain adaptation. And the investigation on the parameters shows that some parameters are important for both the general-domain and in-domain translation and the great change of them during continual training brings about the performance decline in general-domain. We conduct experiments across different language pairs and domains to ensure the validity and reliability of our findings.

Generating Diverse Translation from Model Distribution with Dropout

Despite the improvement of translation quality, neural machine translation (NMT) often suffers from the lack of diversity in its generation. In this paper, we propose to generate diverse translations by deriving a large number of possible models with Bayesian modelling and sampling models from them for inference. The possible models are obtained by applying concrete dropout to the NMT model and each of them has specific confidence for its prediction, which corresponds to a posterior model distribution under specific training data in the principle of Bayesian modeling. With variational inference, the posterior model distribution can be approximated with a variational distribution, from which the final models for inference are sampled. We conducted experiments on Chinese-English and English-German translation tasks and the results shows that our method makes a better trade-off between diversity and accuracy.

Token-level Adaptive Training for Neural Machine Translation

There exists a token imbalance phenomenon in natural language as different tokens appear with different frequencies, which leads to different learning difficulties for tokens in Neural Machine Translation (NMT). The vanilla NMT model usually adopts trivial equal-weighted objectives for target tokens with different frequencies and tends to generate more high-frequency tokens and less low-frequency tokens compared with the golden token distribution. However, low-frequency tokens may carry critical semantic information that will affect the translation quality once they are neglected. In this paper, we explored target token-level adaptive objectives based on token frequencies to assign appropriate weights for each target token during training. We aimed that those meaningful but relatively low-frequency words could be assigned with larger weights in objectives to encourage the model to pay more attention to these tokens. Our method yields consistent improvements in translation quality on ZH-EN, EN-RO, and EN-DE translation tasks, especially on sentences that contain more low-frequency tokens where we can get 1.68, 1.02, and 0.52 BLEU increases compared with baseline, respectively. Further analyses show that our method can also improve the lexical diversity of translation.

Universal Model for Multi-Domain Medical Image Retrieval

Medical Image Retrieval (MIR) helps doctors quickly find similar patients' data, which can considerably aid the diagnosis process. MIR is becoming increasingly helpful due to the wide use of digital imaging modalities and the growth of the medical image repositories. However, the popularity of various digital imaging modalities in hospitals also poses several challenges to MIR. Usually, one image retrieval model is only trained to handle images from one modality or one source. When there are needs to retrieve medical images from several sources or domains, multiple retrieval models need to be maintained, which is cost ineffective. In this paper, we study an important but unexplored task: how to train one MIR model that is applicable to medical images from multiple domains? Simply fusing the training data from multiple domains cannot solve this problem because some domains become over-fit sooner when trained together using existing methods. Therefore, we propose to distill the knowledge in multiple specialist MIR models into a single multi-domain MIR model via universal embedding to solve this problem. Using skin disease, x-ray, and retina image datasets, we validate that our proposed universal model can effectively accomplish multi-domain MIR.

RaSE: Random Subspace Ensemble Classification

We propose a new model-free ensemble classification framework, Random Subspace Ensemble (RaSE), for sparse classification. In the RaSE algorithm, we aggregate many weak learners, where each weak learner is a base classifier trained in a subspace optimally selected from a collection of random subspaces. To conduct subspace selection, we propose a new criterion, ratio information criterion (RIC), based on weighted Kullback-Leibler divergences. The theoretical analysis includes the risk and Monte-Carlo variance of RaSE classifier, establishing the weak consistency of RIC, and providing an upper bound for the misclassification rate of RaSE classifier. An array of simulations under various models and real-data applications demonstrate the effectiveness of the RaSE classifier in terms of low misclassification rate and accurate feature ranking. The RaSE algorithm is implemented in the R package RaSEn on CRAN.