arXiv subjects

Ji Wu

Publications and source records attributed to Ji Wu.

At least 91 records · Page 5Linked to original sources

arXiv2022

Ultra-Low-Frequency Radio Astronomy Observations from a Selenocentric Orbit: first results of the Longjiang-2 experiment

This paper introduces the first results of observations with the Ultra-Long-Wavelength (ULW) -- Low Frequency Interferometer and Spectrometer (LFIS) on board the selenocentric satellite Longjiang-2. We present a brief description of the satellite and focus on the LFIS payload. The in-orbit commissioning confirmed a reliable operational status of the instrumentation. We also present results of a transition observation, which offers unique measurements on several novel aspects. We estimate the RFI suppression required for such a radio astronomy instrumentation at the Moon distances from Earth to be of the order of 80 dB. We analyse a method of separating Earth- and satellite-originated radio frequency interference (RFI). It is found that the RFI level at frequencies lower than a few MHz is smaller than the receiver noise floor.

astro-ph.IM↗

arXiv2022

Understanding the Failure of Batch Normalization for Transformers in NLP

Batch Normalization (BN) is a core and prevalent technique in accelerating the training of deep neural networks and improving the generalization on Computer Vision (CV) tasks. However, it fails to defend its position in Natural Language Processing (NLP), which is dominated by Layer Normalization (LN). In this paper, we are trying to answer why BN usually performs worse than LN in NLP tasks with Transformer models. We find that the inconsistency between training and inference of BN is the leading cause that results in the failure of BN in NLP. We define Training Inference Discrepancy (TID) to quantitatively measure this inconsistency and reveal that TID can indicate BN's performance, supported by extensive experiments, including image classification, neural machine translation, language modeling, sequence labeling, and text classification tasks. We find that BN can obtain much better test performance than LN when TID keeps small through training. To suppress the explosion of TID, we propose Regularized BN (RBN) that adds a simple regularization term to narrow the gap between batch statistics and population statistics of BN. RBN improves the performance of BN consistently and outperforms or is on par with LN on 17 out of 20 settings, involving ten datasets and two common variants of Transformer Our code is available at https://github.com/wjxts/RegularizedBN.

cs.CL↗

arXiv2022

Feature Forgetting in Continual Representation Learning

In continual and lifelong learning, good representation learning can help increase performance and reduce sample complexity when learning new tasks. There is evidence that representations do not suffer from "catastrophic forgetting" even in plain continual learning, but little further fact is known about its characteristics. In this paper, we aim to gain more understanding about representation learning in continual learning, especially on the feature forgetting problem. We devise a protocol for evaluating representation in continual learning, and then use it to present an overview of the basic trends of continual representation learning, showing its consistent deficiency and potential issues. To study the feature forgetting problem, we create a synthetic dataset to identify and visualize the prevalence of feature forgetting in neural networks. Finally, we propose a simple technique using gating adapters to mitigate feature forgetting. We conclude by discussing that improving representation learning benefits both old and new tasks in continual learning.

cs.LG↗

arXiv2022

Table-based Fact Verification with Self-adaptive Mixture of Experts

The table-based fact verification task has recently gained widespread attention and yet remains to be a very challenging problem. It inherently requires informative reasoning over natural language together with different numerical and logical reasoning on tables (e.g., count, superlative, comparative). Considering that, we exploit mixture-of-experts and present in this paper a new method: Self-adaptive Mixture-of-Experts Network (SaMoE). Specifically, we have developed a mixture-of-experts neural network to recognize and execute different types of reasoning -- the network is composed of multiple experts, each handling a specific part of the semantics for reasoning, whereas a management module is applied to decide the contribution of each expert network to the verification result. A self-adaptive method is developed to teach the management module combining results of different experts more efficiently without external knowledge. The experimental results illustrate that our framework achieves 85.1% accuracy on the benchmark dataset TabFact, comparable with the previous state-of-the-art models. We hope our framework can serve as a new baseline for table-based verification. Our code is available at https://github.com/THUMLP/SaMoE.

cs.AI↗

arXiv2021

Distilling effective supervision for robust medical image segmentation with noisy labels

Despite the success of deep learning methods in medical image segmentation tasks, the human-level performance relies on massive training data with high-quality annotations, which are expensive and time-consuming to collect. The fact is that there exist low-quality annotations with label noise, which leads to suboptimal performance of learned models. Two prominent directions for segmentation learning with noisy labels include pixel-wise noise robust training and image-level noise robust training. In this work, we propose a novel framework to address segmenting with noisy labels by distilling effective supervision information from both pixel and image levels. In particular, we explicitly estimate the uncertainty of every pixel as pixel-wise noise estimation, and propose pixel-wise robust learning by using both the original labels and pseudo labels. Furthermore, we present an image-level robust learning method to accommodate more information as the complements to pixel-level learning. We conduct extensive experiments on both simulated and real-world noisy datasets. The results demonstrate the advantageous performance of our method compared to state-of-the-art baselines for medical image segmentation with noisy labels.

cs.CV↗

arXiv2021

Multi-view Graph Contrastive Representation Learning for Drug-Drug Interaction Prediction

Drug-drug interaction(DDI) prediction is an important task in the medical health machine learning community. This study presents a new method, multi-view graph contrastive representation learning for drug-drug interaction prediction, MIRACLE for brevity, to capture inter-view molecule structure and intra-view interactions between molecules simultaneously. MIRACLE treats a DDI network as a multi-view graph where each node in the interaction graph itself is a drug molecular graph instance. We use GCNs and bond-aware attentive message passing networks to encode DDI relationships and drug molecular graphs in the MIRACLE learning stage, respectively. Also, we propose a novel unsupervised contrastive learning component to balance and integrate the multi-view information. Comprehensive experiments on multiple real datasets show that MIRACLE outperforms the state-of-the-art DDI prediction models consistently.

cs.LG↗

arXiv2020

Measuring Information Transfer in Neural Networks

Quantifying the information content in a neural network model is essentially estimating the model's Kolmogorov complexity. Recent success of prequential coding on neural networks points to a promising path of deriving an efficient description length of a model. We propose a practical measure of the generalizable information in a neural network model based on prequential coding, which we term Information Transfer ($L_{IT}$). Theoretically, $L_{IT}$ is an estimation of the generalizable part of a model's information content. In experiments, we show that $L_{IT}$ is consistently correlated with generalizable information and can be used as a measure of patterns or "knowledge" in a model or a dataset. Consequently, $L_{IT}$ can serve as a useful analysis tool in deep learning. In this paper, we apply $L_{IT}$ to compare and dissect information in datasets, evaluate representation models in transfer learning, and analyze catastrophic forgetting and continual learning algorithms. $L_{IT}$ provides an information perspective which helps us discover new insights into neural network learning.

cs.LG↗

arXiv2020

Applications of BERT Based Sequence Tagging Models on Chinese Medical Text Attributes Extraction

We convert the Chinese medical text attributes extraction task into a sequence tagging or machine reading comprehension task. Based on BERT pre-trained models, we have not only tried the widely used LSTM-CRF sequence tagging model, but also other sequence models, such as CNN, UCNN, WaveNet, SelfAttention, etc, which reaches similar performance as LSTM+CRF. This sheds a light on the traditional sequence tagging models. Since the aspect of emphasis for different sequence tagging models varies substantially, ensembling these models adds diversity to the final system. By doing so, our system achieves good performance on the task of Chinese medical text attributes extraction (subtask 2 of CCKS 2019 task 1).

cs.CL↗

arXiv2020

Learning Conceptual-Contextual Embeddings for Medical Text

External knowledge is often useful for natural language understanding tasks. We introduce a contextual text representation model called Conceptual-Contextual (CC) embeddings, which incorporates structured knowledge into text representations. Unlike entity embedding methods, our approach encodes a knowledge graph into a context model. CC embeddings can be easily reused for a wide range of tasks just like pre-trained language models. Our model effectively encodes the huge UMLS database by leveraging semantic generalizability. Experiments on electronic health records (EHRs) and medical text processing benchmarks showed our model gives a major boost to the performance of supervised medical NLP tasks.

cs.CL↗

arXiv2020

Label-guided Learning for Text Classification

Text classification is one of the most important and fundamental tasks in natural language processing. Performance of this task mainly dependents on text representation learning. Currently, most existing learning frameworks mainly focus on encoding local contextual information between words. These methods always neglect to exploit global clues, such as label information, for encoding text information. In this study, we propose a label-guided learning framework LguidedLearn for text representation and classification. Our method is novel but simple that we only insert a label-guided encoding layer into the commonly used text representation learning schemas. That label-guided layer performs label-based attentive encoding to map the universal text embedding (encoded by a contextual information learner) into different label spaces, resulting in label-wise embeddings. In our proposed framework, the label-guided layer can be easily and directly applied with a contextual encoding method to perform jointly learning. Text information is encoded based on both the local contextual information and the global label clues. Therefore, the obtained text embeddings are more robust and discriminative for text classification. Extensive experiments are conducted on benchmark datasets to illustrate the effectiveness of our proposed method.

cs.CL↗

arXiv2020

Tensor Graph Convolutional Networks for Text Classification

Compared to sequential learning models, graph-based neural networks exhibit some excellent properties, such as ability capturing global information. In this paper, we investigate graph-based neural networks for text classification problem. A new framework TensorGCN (tensor graph convolutional networks), is presented for this task. A text graph tensor is firstly constructed to describe semantic, syntactic, and sequential contextual information. Then, two kinds of propagation learning perform on the text graph tensor. The first is intra-graph propagation used for aggregating information from neighborhood nodes in a single graph. The second is inter-graph propagation used for harmonizing heterogeneous information between graphs. Extensive experiments are conducted on benchmark datasets, and the results illustrate the effectiveness of our proposed framework. Our proposed TensorGCN presents an effective way to harmonize and integrate heterogeneous information from different kinds of graphs.

cs.CL↗

arXiv2019

Pick-and-Learn: Automatic Quality Evaluation for Noisy-Labeled Image Segmentation

Deep learning methods have achieved promising performance in many areas, but they are still struggling with noisy-labeled images during the training process. Considering that the annotation quality indispensably relies on great expertise, the problem is even more crucial in the medical image domain. How to eliminate the disturbance from noisy labels for segmentation tasks without further annotations is still a significant challenge. In this paper, we introduce our label quality evaluation strategy for deep neural networks automatically assessing the quality of each label, which is not explicitly provided, and training on clean-annotated ones. We propose a solution for network automatically evaluating the relative quality of the labels in the training set and using good ones to tune the network parameters. We also design an overfitting control module to let the network maximally learn from the precise annotations during the training process. Experiments on the public biomedical image segmentation dataset have proved the method outperforms baseline methods and retains both high accuracy and good generalization at different noise levels.

cs.CV↗

arXiv2019

Discovering the Sky at the Longest Wavelengths with Small Satellite Constellations

Due to ionosphere absorption and the interference by natural and artificial radio emissions, ground observation of the sky at the decameter or longer is very difficult. This unexplored part of electromagnetic spectrum has the potential of great discoveries, notably in the study of cosmic dark ages and dawn, but also in heliophysics and space weather, planets, cosmic ray and neutrinos, pulsar and interstellar medium, extragalactic radio sources, and even SETI. At a forum organized by the International Space Science Institute-Beijing (ISSI-BJ), we discussed the prospect of opening up this window for astronomical observations by using a constellation of small or micro-satellites. We discussed the past experiments and the current ones such as the low frequency payload on Chang'e-4 mission lander, relay satellite and the Longjiang satellite, and also the future DSL mission, which is a linear array on lunar orbit which can make synthesized map of the whole sky as well as measure the global spectrum. We also discuss the synergy with other experiments, including ground global experiments such as EDGES, SARAS, SCI-HI and High-z, PRIZM/Albatros, ground imaging facillities such as LOFAR and MWA, and space experiments such as SUNRISE, DARE/DAPPER and PRATUSH. We also discussed some technical aspects of the DSL concept.

astro-ph.IM↗

arXiv2019

Delta Embedding Learning

Unsupervised word embeddings have become a popular approach of word representation in NLP tasks. However there are limitations to the semantics represented by unsupervised embeddings, and inadequate fine-tuning of embeddings can lead to suboptimal performance. We propose a novel learning technique called Delta Embedding Learning, which can be applied to general NLP tasks to improve performance by optimized tuning of the word embeddings. A structured regularization is applied to the embeddings to ensure they are tuned in an incremental way. As a result, the tuned word embeddings become better word representations by absorbing semantic information from supervision without "forgetting." We apply the method to various NLP tasks and see a consistent improvement in performance. Evaluation also confirms the tuned word embeddings have better semantic properties.

cs.CL↗

arXiv2018

Exploiting Sentence Embedding for Medical Question Answering

Despite the great success of word embedding, sentence embedding remains a not-well-solved problem. In this paper, we present a supervised learning framework to exploit sentence embedding for the medical question answering task. The learning framework consists of two main parts: 1) a sentence embedding producing module, and 2) a scoring module. The former is developed with contextual self-attention and multi-scale techniques to encode a sentence into an embedding tensor. This module is shortly called Contextual self-Attention Multi-scale Sentence Embedding (CAMSE). The latter employs two scoring strategies: Semantic Matching Scoring (SMS) and Semantic Association Scoring (SAS). SMS measures similarity while SAS captures association between sentence pairs: a medical question concatenated with a candidate choice, and a piece of corresponding supportive evidence. The proposed framework is examined by two Medical Question Answering(MedicalQA) datasets which are collected from real-world applications: medical exam and clinical diagnosis based on electronic medical records (EMR). The comparison results show that our proposed framework achieved significant improvements compared to competitive baseline approaches. Additionally, a series of controlled experiments are also conducted to illustrate that the multi-scale strategy and the contextual self-attention layer play important roles for producing effective sentence embedding, and the two kinds of scoring strategies are highly complementary to each other for question answering problems.

cs.CL↗

arXiv2018

Medical Exam Question Answering with Large-scale Reading Comprehension

Reading and understanding text is one important component in computer aided diagnosis in clinical medicine, also being a major research problem in the field of NLP. In this work, we introduce a question-answering task called MedQA to study answering questions in clinical medicine using knowledge in a large-scale document collection. The aim of MedQA is to answer real-world questions with large-scale reading comprehension. We propose our solution SeaReader--a modular end-to-end reading comprehension model based on LSTM networks and dual-path attention architecture. The novel dual-path attention models information flow from two perspectives and has the ability to simultaneously read individual documents and integrate information across multiple documents. In experiments our SeaReader achieved a large increase in accuracy on MedQA over competing models. Additionally, we develop a series of novel techniques to demonstrate the interpretation of the question answering process in SeaReader.

cs.CL↗

arXiv2017

Study on a Poisson's Equation Solver Based On Deep Learning Technique

In this work, we investigated the feasibility of applying deep learning techniques to solve Poisson's equation. A deep convolutional neural network is set up to predict the distribution of electric potential in 2D or 3D cases. With proper training data generated from a finite difference solver, the strong approximation capability of the deep convolutional neural network allows it to make correct prediction given information of the source and distribution of permittivity. With applications of L2 regularization, numerical experiments show that the predication error of 2D cases can reach below 1.5\% and the predication of 3D cases can reach below 3\%, with a significant reduction in CPU time compared with the traditional solver based on finite difference methods.

physics.comp-ph↗

arXiv2017

The Monkeytyping Solution to the YouTube-8M Video Understanding Challenge

This article describes the final solution of team monkeytyping, who finished in second place in the YouTube-8M video understanding challenge. The dataset used in this challenge is a large-scale benchmark for multi-label video classification. We extend the work in [1] and propose several improvements for frame sequence modeling. We propose a network structure called Chaining that can better capture the interactions between labels. Also, we report our approaches in dealing with multi-scale information and attention pooling. In addition, We find that using the output of model ensemble as a side target in training can boost single model performance. We report our experiments in bagging, boosting, cascade, and stacking, and propose a stacking algorithm called attention weighted stacking. Our final submission is an ensemble that consists of 74 sub models, all of which are listed in the appendix.

cs.CV↗