United States

Deep networks are typically trained with many more parameters than the size of the training dataset. Recent empirical evidence indicates that the practice of overparameterization not only benefits training large models, but also assists â€“ perhaps counterintuitively â€“ building lightweight models. Specifically, it suggests that overparameterization benefits model pruning / sparsification. This paper sheds light on these empirical findings by theoretically characterizing the high-dimensional asymptotics of model pruning in the overparameterized regime. The theory presented addresses the following core question: ``should one train a small model from the beginning, or first train a large model and then prune?&#39;&#39;. We analytically identify regimes in which, even if the location of the most informative features is known, we are better off fitting a large model and then pruning rather than simply training with the known informative features. This leads to a new double descent in the training of sparse models: growing the original model, while preserving the target sparsity, improves the test accuracy as one moves beyond the overparameterization threshold. Our analysis further reveals the benefit of retraining by relating it to feature correlations. We find that the above phenomena are already present in linear and random-features models. Our technical approach advances the toolset of high-dimensional analysis and precisely characterizes the asymptotic distribution of over-parameterized least-squares. The intuition gained by analytically studying simpler models is numerically verified on neural networks.

AAAI 2021

Provable Benefits of Overparameterization in Model Compression: From Double Descent to Pruning Neural Networks

**Welcome to EMNLP 2023!**

On behalf of the EMNLP 2023 Organizing Committee, I extend a warm and heartfelt welcome to all of you to the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP). It is with immense pleasure and excitement that we gather here in Singapore, a vibrant hub of innovation and technological advancement.

The conference program is packed with insightful presentations, thought-provoking workshops, and engaging networking opportunities. In addition to the technical sessions, we have also planned several social events that will provide you with the opportunity to connect with your colleagues.

The conference is held in person at the Resorts World Convention Centre in Singapore, and available on-line with the help of Underline.

For those in Singapore, I hope that you will find time to explore the exciting Sentosa Island, Gardens by the Bay, Singapore Botanic Gardens and the many other attractions unique to Singapore.

Of course, we are grateful to our sponsors and partners for their generous support of EMNLP 2023, whose contributions make it possible for us to host this world-class event.

Yuji Matsumoto (RIKEN AIP) 
EMNLP 2023 General Chair

To access the **EMNLP 2023** event page on Underline, you need to register for the Conference. 
Please follow **[this link](https://2023.emnlp.org/registration/)** for more details.

EMNLP 2023

EMNLP 2023 took place in Singapore from Dec 6th to Dec 10th, 2023.

Deep networks are typically trained with many more parameters than the size of the training dataset. Recent empirical evidence indicates that the practice of overparameterization not only benefits training large models, but also assists â€“ perhaps counterintuitively â€“ building lightweight models. Specifically, it suggests that overparameterization benefits model pruning / sparsification. This paper sheds light on these empirical findings by theoretically characterizing the high-dimensional asymptotics of model pruning in the overparameterized regime. The theory presented addresses the following core question: ``should one train a small model from the beginning, or first train a large model and then prune?''. We analytically identify regimes in which, even if the location of the most informative features is known, we are better off fitting a large model and then pruning rather than simply training with the known informative features. This leads to a new double descent in the training of sparse models: growing the original model, while preserving the target sparsity, improves the test accuracy as one moves beyond the overparameterization threshold. Our analysis further reveals the benefit of retraining by relating it to feature correlations. We find that the above phenomena are already present in linear and random-features models. Our technical approach advances the toolset of high-dimensional analysis and precisely characterizes the asymptotic distribution of over-parameterized least-squares. The intuition gained by analytically studying simpler models is numerically verified on neural networks.

technical paper

The purpose of the AAAI conference is to promote research in artificial intelligence (AI) and scientific exchange among AI researchers, practitioners, scientists, and engineers in affiliated disciplines. 

Unsupervised anomaly detection aims to identify data samples that have low probability density from a set of input samples, and only the normal samples are provided for model training. The inference of abnormal regions on the input image requires an understanding of the surrounding semantic context. This work presents a Semantic Context based Anomaly Detection Network, SCADN, for unsupervised anomaly detection by learning the semantic context from the normal samples. To achieve this, we first generate multi-scale striped masks to remove a part of regions  from the normal samples, and then train a generative adversarial network to reconstruct the unseen regions. Note that the masks are designed in multiple scales and stripe directions, and various training examples are generated to obtain the rich semantic context . In testing, we obtain an error map by computing the difference between the reconstructed image and the input image for all samples, and infer the abnormal samples based on the error maps. Finally, we perform various experiments on three public benchmark datasets and a new dataset LaceAD collected by us, and show that our method clearly outperforms the current state-of-the-art methods.

Learning Semantic Context from Normal Samples for Unsupervised Anomaly Detection

The immunohistochemistry (IHC) test of biopsy tissue is crucial to develop targeted treatment and evaluate prognosis for cancer patients. The IHC staining slide is usually digitized into the whole-slide image (WSI) with gigapixels for quantitative image analysis. To perform a whole image prediction (e.g., IHC scoring, survival prediction, and cancer grading) from this kind of high-dimensional image, algorithms are often developed based on multi-instance learning (MIL) framework. However, the multi-scale information of WSI and the associations among instances are not well explored in existing MIL based studies. Inspired by the fact that pathologists jointly analyze visual fields at multiple powers of objective for diagnostic predictions, we propose a Pathologist-Tree Network (PTree-Net) to sparsely model the WSI efficiently in multi-scale manner. Specifically, we propose a Focal-Aware Module (FAM) that can approximately estimate diagnosis-related regions with an extractor trained using the thumbnail of WSI. With the initial diagnosis-related regions, we hierarchically model the multi-scale patches in a tree structure, where both the global and local information can be captured. To explore this tree structure in an end-to-end network, we propose a patch Relevance-enhanced Graph Convolutional Network (RGCN) to explicitly model the correlations of adjacent parent-child nodes, accompanied by patch relevance to exploit the implicit contextual information among distant nodes. In addition, tree-based self-supervision is devised to improve representation learning and suppress irrelevant instances adaptively. Extensive experiments are performed on a large-scale IHC HER2 dataset. The ablation study confirms the effectiveness of our design, and our approach outperforms state-of-the-art by a large margin.

Diagnose Like a Pathologist: Weakly-Supervised Pathologist-Tree Network for Slide-Level Immunohistochemical Scoring

Randomized algorithms are used in many state-of-the-art solvers for constraint satisfaction problems (CSP) and Boolean satisfiability (SAT) problems. For many of these problems, there is no single solver which will dominate others. Having access to the underlying runtime distributions (RTD) of these solvers can allow for better use of algorithm selection, algorithm portfolios, and restart strategies. Previous state-of-the-art methods directly try to predict a fixed parametric distribution that the input instance follows. In this paper, we extend RTD prediction models into the Bayesian setting for the first time. This new model achieves robust predictive performance in the low observation setting, as well as handling censored observations. This technique also allows for richer representations which cannot be achieved by the classical models which restrict their output representations. Our model outperforms the previous state-of-the-art model in settings in which data is scarce, and can make use of censored data such as lower bound time estimates, where that type of data would otherwise be discarded. It can also quantify its uncertainty in its predictions, allowing for algorithm portfolio models to make better informed decisions about which algorithm to run on a particular instance.

Bayes DistNet - A Robust Neural Network for Algorithm Runtime Distribution Predictions

In this paper, we study the problem of embedding uncertain knowledge graphs, where each relation between entities is associated with a confidence score. Observing the existing embedding methods may discard the uncertainty information, only incorporate a specific type of score function, or cause many false-negative samples in the training, we propose the PASSLEAF framework to solve the above issues. PASSLEAF consists of two parts, one is a model that can incorporate different types of scoring functions to predict the relation confidence scores and the other is the semi-supervised learning model by exploiting both positive and negative samples associated with the estimated confidence scores. Furthermore, PASSLEAF leverages a sample pool as a relay of generated samples to further augment the semi-supervised learning. Experiment results show that our proposed framework can learn better embedding in terms of having higher accuracy in both the confidence score prediction and tail entity prediction.

PASSLEAF: A Pool-Based Semi-Supervised Learning Framework for Uncertain Knowledge Graph Embedding

In the fight against the COVID-19 pandemic, many social activities have moved online; society's overwhelming reliance on the complex cyberspace makes its security more important than ever. In this paper, we propose and develop an intelligent system named Dr.HIN to protect users against the evolving Android malware attacks in the COVID-19 era and beyond. In Dr.HIN, besides app content, we propose to consider higher-level semantics and social relations among apps, developers and mobile devices to comprehensively depict Android apps; and then we introduce a structured heterogeneous information network (HIN) to model the complex relations and exploit meta-path guided strategy to learn node (i.e., app) representations from HIN. As the representations of malware could be highly entangled with benign apps in the complex ecosystem of development, it poses a new challenge of learning the latent explanatory factors hidden in the HIN embeddings to detect the evolving malware. To address this challenge, we propose to integrate domain priors generated from different views (i.e., app content, app authorship, app installation) to devise an adversarial disentangler to separate the distinct, informative factors of variations hidden in the HIN embeddings for large-scale Android malware detection. This is the first attempt of disentangled representation learning in HIN data. Promising experimental results based on the large-scale and real sample collections from security industry demonstrate the performance of Dr.HIN in evolving Android malware detection, by comparison with baselines and popular mobile security products.

Disentangled Representation Learning in Heterogeneous Information Network for Large-Scale Android Malware Detection in the COVID-19 Era and Beyond

Extreme multi-label text classification(XMC) is a task for finding the most relevant labels from a large label set. Nowadays deep learning-based methods have shown significant success in XMC. However, the existing methods (e.g., AttentionXML and X-Transformer etc) still suffer from 1) combining several models to train and predict for one dataset, and 2) sampling negative labels statically during the process of training label ranking model, which will harm the performance and accuracy of model. To address the above problems, we propose LightXML, which adopts end-to-end training and dynamical negative labels sampling. In LightXML, we use GAN like networks to recall and rank labels. The label recalling part will generate negative and positive labels, and the label ranking part will distinguish positive labels from these labels. Based on these networks, negative labels are sampled dynamically during label ranking part training. With feeding both label recalling and ranking parts with the same text representation, LightXML can reach high performance. Extensive experiments show that LightXML outperforms state-of-the-art methods in five extreme multi-label datasets with much smaller model size and lower computational complexity. In particular, on the Amazon dataset with  670K labels, LightXML can reduce the model size up to 72% compared to AttentionXML. Our code is available at http://github.com/kongds/LightXML.

LightXML: Transformer with Dynamic Negative Sampling for High-Performance Extreme Multi-Label Text Classification

One important challenge of applying deep learning to electronic health records (EHR) is the complexity of their multimodal structure. EHR usually contains a mixture of structured (codes) and unstructured (free-text) data with sparse and irregular longitudinal features -- all of which doctors utilize when making decisions. In the deep learning regime, determining how different modality representations should be fused together is a difficult problem, which is often addressed by handcrafted modeling and intuition. In this work, we extend state-of-the-art neural architecture search (NAS) methods and propose MUltimodal Fusion Architecture SeArch (MUFASA) to simultaneously search across multimodal fusion strategies and modality-specific architectures for the first time. We demonstrate empirically that our MUFASA method outperforms established unimodal NAS on public EHR data with comparable computation costs. In addition, MUFASA produces architectures that outperform Transformer and Evolved Transformer. Compared with these baselines on CCS diagnosis code prediction, our discovered models improve top-5 recall from 0.88 to 0.91 and demonstrate the ability to generalize to other EHR tasks. Studying our top architecture in depth, we provide empirical evidence that MUFASA's improvements are derived from its ability to both customize modeling for each modality and find effective fusion strategies.

MUFASA: Multimodal Fusion Architecture Search for Electronic Health Records

Recent studies on machine reading comprehension have focused on text-level understanding but have not yet reached the level of human understanding of the visual layout and content of real-world documents. In this study, we introduce a new visual machine reading comprehension dataset, named VisualMRC, wherein given a question and a document image, a machine reads and comprehends texts in the image to answer the question in natural language. Compared with existing visual question answering datasets that contain texts in images, VisualMRC focuses more on developing natural language understanding and generation abilities. It contains 30,000+ pairs of a question and an abstractive answer for 10,000+ document images sourced from multiple domains of webpages. We also introduce a new model that extends existing sequence-to-sequence models, pre-trained with large-scale text corpora, to take into account the visual layout and content of documents. Experiments with VisualMRC show that this model outperformed the base sequence-to-sequence models and a state-of-the-art VQA model. However, its performance is still below that of humans on most automatic evaluation metrics.  The dataset will facilitate research aimed at connecting vision and language understanding.

VisualMRC: Machine Reading Comprehension on Document Images

Given limited handwriting scripts, humans can easily visualize (or imagine) what the handwritten words/texts would look like with other arbitrary textual contents. Moreover, a person also is able to imitate the handwriting styles of provided reference samples. Humans can do such hallucinations, perhaps because they can learn to disentangle the calligraphic styles and textual contents from given handwriting scripts. However, computers cannot study to do such flexible handwriting imitation with existing techniques. In this paper, we propose a novel handwriting imitation generative adversarial network (HiGAN) to mimic such hallucinations. Specifically, HiGAN can generate variable-length handwritten words/texts conditioned on arbitrary textual contents, which are unconstrained to any predefined corpus or out-of-vocabulary words. Moreover, HiGAN can flexibly control the handwriting styles of synthetic images by disentangling calligraphic styles from the reference samples. Experiments on handwriting benchmarks validate our superiority in terms of visual quality and scalability when comparing to the state-of-the-art methods for handwritten word/text synthesis. The code and pre-trained models can be found at https://github.com/ganji15/HiGAN.

HiGAN: Handwriting Imitation Conditioned on Arbitrary-Length Texts and Disentangled Styles

In wearable-sensor-based activity recognition, it is often assumed that the training and test samples follow the same data distribution. This assumption neglects practical scenarios where the activity patterns inevitably vary from person to person. To solve this problem, transfer learning and domain adaptation approaches are often leveraged to reduce the gaps between different participants. Nevertheless, these approaches require additional information (i.e., labeled or unlabeled data, meta-information) from the target domain during the training stage. In this paper, we introduce a novel method named Generalizable Independent Latent Excitation (GILE) for human activity recognition, which greatly enhances the cross-person generalization capability of the model. Our proposed method is superior to existing methods in the sense that it does not require any access to the target domain information. Besides, this novel model can be directly applied to various target domains without re-training or fine-tuning. Specifically, the proposed model learns to automatically disentangle domain-agnostic and domain-specific features, the former of which are expected to be invariant across various persons. To further remove correlations between the two types of features, a novel Independent Excitation mechanism is incorporated in the latent feature space. Comprehensive experimental evaluations are conducted on three benchmark datasets to demonstrate the superiority of the proposed method over the state-of-the-art solutions.

Downloads

Next from AAAI 2021

Learning Semantic Context from Normal Samples for Unsupervised Anomaly Detection

Similar lecture

Data Pruning for Efficient Model Pruning in Neural Machine Translation

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAAI 2021

Learning Semantic Context from Normal Samples for Unsupervised Anomaly Detection

Similar lecture

Data Pruning for Efficient Model Pruning in Neural Machine Translation

Downloads