Search

Low sky brightness is crucial for ground-based astronomical observations, because it limits the observational capability to detect fainter sources. Lenghu, located on the Tibetan Plateau in China, has been identified as an high-quality astronomical site in China, including dark sky in optical band. In this work, we will report the preliminary results of near-infrared sky brightness measurements at Lenghu. Utilizing a wide-field small telescope equipped with an InGaAs camera, we have been conducting long-term monitoring of near-infrared sky brightness in the J and H' bands, respectively, since January 2024. For each image, photometry and astrometry were performed, then sky background was calibrated by standard stars from the 2MASS catalog. This report includes preliminary results on the sky brightness at zenith in the J and H' bands, as well as their variations with solar elevation at Lenghu. Our initial results indicate that the near-infrared sky brightness at Lenghu is comparable to that of other world-class sites, and long-term monitoring will be continued., Comment: 8 pages, 8 figures, presented at 2024 SPIE Astronomical Telescopes + Instrumentation conference

Recently, intelligent analysis of lung nodules with the assistant of computer aided detection (CAD) techniques can improve the accuracy rate of lung cancer diagnosis. However, existing CAD systems and pulmonary datasets mainly focus on Computed Tomography (CT) images from one single period, while ignoring the cross spatio-temporal features associated with the progression of nodules contained in imaging data from various captured periods of lung cancer. If the evolution patterns of nodules across various periods in the patients' CT sequences can be explored, it will play a crucial role in guiding the precise screening identification of lung cancer. Therefore, a cross spatio-temporal lung nodule dataset with pathological information for nodule identification and diagnosis is constructed, which contains 328 CT sequences and 362 annotated nodules from 109 patients. This comprehensive database is intended to drive research in the field of CAD towards more practical and robust methods, and also contribute to the further exploration of precision medicine related field. To ensure patient confidentiality, we have removed sensitive information from the dataset.

This work is motivated by two key trends. On one hand, large language models (LLMs) have shown remarkable versatility in various generative tasks such as writing, drawing, and question answering, significantly reducing the time required for many routine tasks. On the other hand, researchers, whose work is not only time-consuming but also highly expertise-demanding, face increasing challenges as they have to spend more time reading, writing, and reviewing papers. This raises the question: how can LLMs potentially assist researchers in alleviating their heavy workload? This study focuses on the topic of LLMs assist NLP Researchers, particularly examining the effectiveness of LLM in assisting paper (meta-)reviewing and its recognizability. To address this, we constructed the ReviewCritique dataset, which includes two types of information: (i) NLP papers (initial submissions rather than camera-ready) with both human-written and LLM-generated reviews, and (ii) each review comes with "deficiency" labels and corresponding explanations for individual segments, annotated by experts. Using ReviewCritique, this study explores two threads of research questions: (i) "LLMs as Reviewers", how do reviews generated by LLMs compare with those written by humans in terms of quality and distinguishability? (ii) "LLMs as Metareviewers", how effectively can LLMs identify potential issues, such as Deficient or unprofessional review segments, within individual paper reviews? To our knowledge, this is the first work to provide such a comprehensive analysis.

Supervised learning methods have been found to exhibit inductive biases favoring simpler features. When such features are spuriously correlated with the label, this can result in suboptimal performance on minority subgroups. Despite the growing popularity of methods which learn from unlabeled data, the extent to which these representations rely on spurious features for prediction is unclear. In this work, we explore the impact of spurious features on Self-Supervised Learning (SSL) for visual representation learning. We first empirically show that commonly used augmentations in SSL can cause undesired invariances in the image space, and illustrate this with a simple example. We further show that classical approaches in combating spurious correlations, such as dataset re-sampling during SSL, do not consistently lead to invariant representations. Motivated by these findings, we propose LateTVG to remove spurious information from these representations during pre-training, by regularizing later layers of the encoder via pruning. We find that our method produces representations which outperform the baselines on several benchmarks, without the need for group or label information during SSL.

In recent years, the buffer layer ablation failures of high voltage cables are frequently reported by the power systems. Previous studies have dominantly regarded the buffer layer as the continuous homogeneous medium, whereas neglects its microstructures. In this paper, the current distribution within the random fiber networks of buffer layer are investigated. Experiment results of our self-designed platform revealed an uneven current distribution in buffer layer at the moment of bearing current. This phenomenon is named as the intrinsic current concentration where the current density concentrates at certain sites inner the buffer layer. And the degree of current concentration will be suppressed by compressing the sample. Then, a 2D simulation model of the random fiber networks was constructed based on the Mikado model. The simulation results also presented an uneven current distribution in the networks whose every fiber can be viewed as a micro-resistor. Two types of dimensionless current concentration factors were defined to describe the degree of current concentration, finding their values decreasing with the rise of fiber density. Meanwhile, it is equivalent of compressing the buffer layer and increasing the fiber density of model. We believe that the intrinsic current concentration phenomenon is mainly related with the inhomogeneity of geometry structure of buffer layer. The ablation traces and fracture fibers observed by the X-ray micro-computed tomography test supported this point. In addition, the non-ideal surface of sample can also induce this phenomenon. The intrinsic current concentration can aggravate the degree of originally existed macroscopic current concentration in cables, thus causing the ablation failure. Our work may unveil a deeper understanding on the cable ablation failure and the electrical response of the similar fibrous materials., Comment: 8 pages, 12 figures

The use of multimodal data in assisted diagnosis and segmentation has emerged as a prominent area of interest in current research. However, one of the primary challenges is how to effectively fuse multimodal features. Most of the current approaches focus on the integration of multimodal features while ignoring the correlation and consistency between different modal features, leading to the inclusion of potentially irrelevant information. To address this issue, we introduce an innovative Multimodal Information Cross Transformer (MicFormer), which employs a dual-stream architecture to simultaneously extract features from each modality. Leveraging the Cross Transformer, it queries features from one modality and retrieves corresponding responses from another, facilitating effective communication between bimodal features. Additionally, we incorporate a deformable Transformer architecture to expand the search space. We conducted experiments on the MM-WHS dataset, and in the CT-MRI multimodal image segmentation task, we successfully improved the whole-heart segmentation DICE score to 85.57 and MIoU to 75.51. Compared to other multimodal segmentation techniques, our method outperforms by margins of 2.83 and 4.23, respectively. This demonstrates the efficacy of MicFormer in integrating relevant information between different modalities in multimodal tasks. These findings hold significant implications for multimodal image tasks, and we believe that MicFormer possesses extensive potential for broader applications across various domains. Access to our method is available at https://github.com/fxxJuses/MICFormer

Federated learning (FL) enables collaborative learning across multiple clients. In most FL work, all clients train a single learning task. However, the recent proliferation of FL applications may increasingly require multiple FL tasks to be trained simultaneously, sharing clients' computing and communication resources, which we call Multiple-Model Federated Learning (MMFL). Current MMFL algorithms use naive average-based client-task allocation schemes that can lead to unfair performance when FL tasks have heterogeneous difficulty levels, e.g., tasks with larger models may need more rounds and data to train. Just as naively allocating resources to generic computing jobs with heterogeneous resource needs can lead to unfair outcomes, naive allocation of clients to FL tasks can lead to unfairness, with some tasks having excessively long training times, or lower converged accuracies. Furthermore, in the FL setting, since clients are typically not paid for their training effort, we face a further challenge that some clients may not even be willing to train some tasks, e.g., due to high computational costs, which may exacerbate unfairness in training outcomes across tasks. We address both challenges by firstly designing FedFairMMFL, a difficulty-aware algorithm that dynamically allocates clients to tasks in each training round. We provide guarantees on airness and FedFairMMFL's convergence rate. We then propose a novel auction design that incentivizes clients to train multiple tasks, so as to fairly distribute clients' training efforts across the tasks. We show how our fairness-based learning and incentive mechanisms impact training convergence and finally evaluate our algorithm with multiple sets of learning tasks on real world datasets.

Enhancing the energy efficiency of buildings significantly relies on monitoring indoor ambient temperature. The potential limitations of conventional temperature measurement techniques, together with the omnipresence of smartphones, have redirected researchers'attention towards the exploration of phone-based ambient temperature estimation methods. However, existing phone-based methods face challenges such as insufficient privacy protection, difficulty in adapting models to various phones, and hurdles in obtaining enough labeled training data. In this study, we propose a distributed phone-based ambient temperature estimation system which enables collaboration among multiple phones to accurately measure the ambient temperature in different areas of an indoor space. This system also provides an efficient, cost-effective approach with a few-shot meta-learning module and an automated label generation module. It shows that with just 5 new training data points, the temperature estimation model can adapt to a new phone and reach a good performance. Moreover, the system uses crowdsourcing to generate accurate labels for all newly collected training data, significantly reducing costs. Additionally, we highlight the potential of incorporating federated learning into our system to enhance privacy protection. We believe this study can advance the practical application of phone-based ambient temperature measurement, facilitating energy-saving efforts in buildings.

Uncovering early-stage metrics that reflect final model performance is one core principle for large-scale pretraining. The existing scaling law demonstrates the power-law correlation between pretraining loss and training flops, which serves as an important indicator of the current training state for large language models. However, this principle only focuses on the model's compression properties on the training data, resulting in an inconsistency with the ability improvements on the downstream tasks. Some follow-up works attempted to extend the scaling-law to more complex metrics (such as hyperparameters), but still lacked a comprehensive analysis of the dynamic differences among various capabilities during pretraining. To address the aforementioned limitations, this paper undertakes a comprehensive comparison of model capabilities at various pretraining intermediate checkpoints. Through this analysis, we confirm that specific downstream metrics exhibit similar training dynamics across models of different sizes, up to 67 billion parameters. In addition to our core findings, we've reproduced Amber and OpenLLaMA, releasing their intermediate checkpoints. This initiative offers valuable resources to the research community and facilitates the verification and exploration of LLM pretraining by open-source researchers. Besides, we provide empirical summaries, including performance comparisons of different models and capabilities, and tuition of key metrics for different training phases. Based on these findings, we provide a more user-friendly strategy for evaluating the optimization state, offering guidance for establishing a stable pretraining process.

Let $\mathcal {F}$ be a given family of graphs. A graph $G$ is $\mathcal {F}$-free if it does not contain any member of $\mathcal {F}$ as a subgraph. Let $C_{l, l}$ be a graph obtained from $2C_l$ such that the two cycles share a common vertex, where $l\geqslant3 $. A $\mathrm{Theta}$ graph is obtained from a cycle $C_k$ by adding an additional edge between two non-consecutive vertices on $C_k$, where $k\geqslant 4$. Let $\Theta_k$ be the set of $\mathrm{Theta}$ graphs on $k$ vertices, where $k \geqslant 4$. For sufficiently large $n $, the unique extremal planar graph with the maximum spectral radius among $C_{l, l}$-free planar graphs on $n$ vertices and among $\Theta_k$-free planar graphs on $n$ vertices are characterized respectively, where $l \geqslant 3$ and $k \geqslant 4$.

The third ML4H symposium was held in person on December 10, 2023, in New Orleans, Louisiana, USA. The symposium included research roundtable sessions to foster discussions between participants and senior researchers on timely and relevant topics for the \ac{ML4H} community. Encouraged by the successful virtual roundtables in the previous year, we organized eleven in-person roundtables and four virtual roundtables at ML4H 2022. The organization of the research roundtables at the conference involved 17 Senior Chairs and 19 Junior Chairs across 11 tables. Each roundtable session included invited senior chairs (with substantial experience in the field), junior chairs (responsible for facilitating the discussion), and attendees from diverse backgrounds with interest in the session's topic. Herein we detail the organization process and compile takeaways from these roundtable discussions, including recent advances, applications, and open challenges for each topic. We conclude with a summary and lessons learned across all roundtables. This document serves as a comprehensive review paper, summarizing the recent advancements in machine learning for healthcare as contributed by foremost researchers in the field., Comment: ML4H 2023, Research Roundtables

Recent studies have demonstrated remarkable performance in time series forecasting. However, due to the partially-observed nature of real-world applications, solely focusing on the target of interest, so-called endogenous variables, is usually insufficient to guarantee accurate forecasting. Notably, a system is often recorded into multiple variables, where the exogenous series can provide valuable external information for endogenous variables. Thus, unlike prior well-established multivariate or univariate forecasting that either treats all the variables equally or overlooks exogenous information, this paper focuses on a practical setting, which is time series forecasting with exogenous variables. We propose a novel framework, TimeXer, to utilize external information to enhance the forecasting of endogenous variables. With a deftly designed embedding layer, TimeXer empowers the canonical Transformer architecture with the ability to reconcile endogenous and exogenous information, where patch-wise self-attention and variate-wise cross-attention are employed. Moreover, a global endogenous variate token is adopted to effectively bridge the exogenous series into endogenous temporal patches. Experimentally, TimeXer significantly improves time series forecasting with exogenous variables and achieves consistent state-of-the-art performance in twelve real-world forecasting benchmarks.

Tokenization is a foundational step in Natural Language Processing (NLP) tasks, bridging raw text and language models. Existing tokenization approaches like Byte-Pair Encoding (BPE) originate from the field of data compression, and it has been suggested that the effectiveness of BPE stems from its ability to condense text into a relatively small number of tokens. We test the hypothesis that fewer tokens lead to better downstream performance by introducing PathPiece, a new tokenizer that segments a document's text into the minimum number of tokens for a given vocabulary. Through extensive experimentation we find this hypothesis not to be the case, casting doubt on the understanding of the reasons for effective tokenization. To examine which other factors play a role, we evaluate design decisions across all three phases of tokenization: pre-tokenization, vocabulary construction, and segmentation, offering new insights into the design of effective tokenizers. Specifically, we illustrate the importance of pre-tokenization and the benefits of using BPE to initialize vocabulary construction. We train 64 language models with varying tokenization, ranging in size from 350M to 2.4B parameters, all of which are made publicly available.

Although several efforts have been taken to enhance the photonic spin Hall shift in deep-subwavelength region, according to effective medium theory, the fundamental confliction between near-zero reflection coefficient and near-zero incident angle still hinders the further application. Here, we reveal a fundamental breakdown of effective medium theory due to the existing of topological singularity in deep-subwavelength region in one-dimensional photonic crystals. We find that near the topological singularity, huge photonic spin Hall shift can be achieved for s-polarization and p-polarization. At the topological singularity, the reflected filed is split as dipole-like distribution with zero photonic spin Hall shift for both-polarizations, which is resulted from the interfere of the spin-maintained normal light and spin-flipped abnormal light. Based on the theoretical research, dual-polarizations thickness and dielectric constant sensing devices can be designed in deep-subwavelength region. Further more, by applying more complicated layered structure, multi-channels dual-polarizations detection and broadband dual-polarizations huge spin Hall shift platform can be designed. This work paves the way to exploring the topological properties and polarization control of photonic crystals and provides a prospective method for the design of multi-channels sensitive detection spin optical devices.

Deep learning has contributed remarkably to the advancement of time series analysis. Still, deep models can encounter performance bottlenecks in real-world data-scarce scenarios, which can be concealed due to the performance saturation with small models on current benchmarks. Meanwhile, large models have demonstrated great powers in these scenarios through large-scale pre-training. Continuous progress has been achieved with the emergence of large language models, exhibiting unprecedented abilities such as few-shot generalization, scalability, and task generality, which are however absent in small deep models. To change the status quo of training scenario-specific small models from scratch, this paper aims at the early development of large time series models (LTSM). During pre-training, we curate large-scale datasets with up to 1 billion time points, unify heterogeneous time series into single-series sequence (S3) format, and develop the GPT-style architecture toward LTSMs. To meet diverse application needs, we convert forecasting, imputation, and anomaly detection of time series into a unified generative task. The outcome of this study is a Time Series Transformer (Timer), which is generative pre-trained by next token prediction and adapted to various downstream tasks with promising capabilities as an LTSM. Code and datasets are available at: https://github.com/thuml/Large-Time-Series-Model.

This paper provides a comprehensive review of the latest advancements in fetal motion correction in MRI. We delve into various contemporary methodologies and technological advancements aimed at overcoming these challenges. It includes traditional 3D fetal MRI correction methods like Slice to Volume Registration (SVR), deep learning-based techniques such as Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) Networks, Transformers, Generative Adversarial Networks (GANs) and most recent advancements of Diffusion Models. The insights derived from this literature review reflect a thorough understanding of both the technical intricacies and practical implications of fetal motion in MRI studies, offering a reasoned perspective on potential solutions and future improvements in this field., Comment: 8 pages, 2 figures

Multi-modal information retrieval (MMIR) is a rapidly evolving field, where significant progress, particularly in image-text pairing, has been made through advanced representation learning and cross-modality alignment research. However, current benchmarks for evaluating MMIR performance in image-text pairing within the scientific domain show a notable gap, where chart and table images described in scholarly language usually do not play a significant role. To bridge this gap, we develop a specialised scientific MMIR (SciMMIR) benchmark by leveraging open-access paper collections to extract data relevant to the scientific domain. This benchmark comprises 530K meticulously curated image-text pairs, extracted from figures and tables with detailed captions in scientific documents. We further annotate the image-text pairs with two-level subset-subcategory hierarchy annotations to facilitate a more comprehensive evaluation of the baselines. We conducted zero-shot and fine-tuning evaluations on prominent multi-modal image-captioning and visual language models, such as CLIP and BLIP. Our analysis offers critical insights for MMIR in the scientific domain, including the impact of pre-training and fine-tuning settings and the influence of the visual and textual encoders. All our data and checkpoints are publicly available at https://github.com/Wusiwei0410/SciMMIR., Comment: camera-ready version for ACL 2024 Findings

As the capabilities of large multimodal models (LMMs) continue to advance, evaluating the performance of LMMs emerges as an increasing need. Additionally, there is an even larger gap in evaluating the advanced knowledge and reasoning abilities of LMMs in non-English contexts such as Chinese. We introduce CMMMU, a new Chinese Massive Multi-discipline Multimodal Understanding benchmark designed to evaluate LMMs on tasks demanding college-level subject knowledge and deliberate reasoning in a Chinese context. CMMMU is inspired by and strictly follows the annotation and analysis pattern of MMMU. CMMMU includes 12k manually collected multimodal questions from college exams, quizzes, and textbooks, covering six core disciplines: Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, and Tech & Engineering, like its companion, MMMU. These questions span 30 subjects and comprise 39 highly heterogeneous image types, such as charts, diagrams, maps, tables, music sheets, and chemical structures. CMMMU focuses on complex perception and reasoning with domain-specific knowledge in the Chinese context. We evaluate 11 open-source LLMs and one proprietary GPT-4V(ision). Even GPT-4V only achieves accuracies of 42%, indicating a large space for improvement. CMMMU will boost the community to build the next-generation LMMs towards expert artificial intelligence and promote the democratization of LMMs by providing diverse language contexts.

In machine learning (ML), a widespread adage is that the area under the precision-recall curve (AUPRC) is a superior metric for model comparison to the area under the receiver operating characteristic (AUROC) for binary classification tasks with class imbalance. This paper challenges this notion through novel mathematical analysis, illustrating that AUROC and AUPRC can be concisely related in probabilistic terms. We demonstrate that AUPRC, contrary to popular belief, is not superior in cases of class imbalance and might even be a harmful metric, given its inclination to unduly favor model improvements in subpopulations with more frequent positive labels. This bias can inadvertently heighten algorithmic disparities. Prompted by these insights, a thorough review of existing ML literature was conducted, utilizing large language models to analyze over 1.5 million papers from arXiv. Our investigation focused on the prevalence and substantiation of the purported AUPRC superiority. The results expose a significant deficit in empirical backing and a trend of misattributions that have fuelled the widespread acceptance of AUPRC's supposed advantages. Our findings represent a dual contribution: a significant technical advancement in understanding metric behaviors and a stark warning about unchecked assumptions in the ML community. All experiments are accessible at https://github.com/mmcdermott/AUC_is_all_you_need.

The importance of personal mobility data is widely recognized in various fields. However, the utilization of real personal mobility data raises privacy concerns. Therefore, it is crucial to generate pseudo personal mobility data that accurately reflects real-world mobility patterns while safeguarding user privacy. Nevertheless, existing methods for generating pseudo mobility data, such as mechanism-based and deep-learning-based approaches, have limitations in capturing sufficient individual heterogeneity. To address these gaps, taking pseudo-person(avatar) as ground-zero, a novel individual-based human mobility generator called GeoAvatar has been proposed - which considers individual heterogeneity in spatial and temporal decision-making, incorporates demographic characteristics, and provides interpretability. Our method utilizes a deep generative model to simulate heterogeneous individual life patterns, a reliable labeler for inferring individual demographic characteristics, and a Bayesian approach for generating spatial choices. Through our method, we have achieved the generation of heterogeneous individual human mobility data without accessing individual-level personal information, with good quality - we evaluated the proposed method based on physical features, activity patterns, and spatial-temporal characteristics, demonstrating its good performance, compared to mechanism-based modeling and black-box deep learning approaches. Furthermore, this method maintains extensibility for broader applications, making it a promising paradigm for generating human mobility data.

As artificial intelligence (AI) rapidly approaches human-level performance in medical imaging, it is crucial that it does not exacerbate or propagate healthcare disparities. Prior research has established AI's capacity to infer demographic data from chest X-rays, leading to a key concern: do models using demographic shortcuts have unfair predictions across subpopulations? In this study, we conduct a thorough investigation into the extent to which medical AI utilizes demographic encodings, focusing on potential fairness discrepancies within both in-distribution training sets and external test sets. Our analysis covers three key medical imaging disciplines: radiology, dermatology, and ophthalmology, and incorporates data from six global chest X-ray datasets. We confirm that medical imaging AI leverages demographic shortcuts in disease classification. While correcting shortcuts algorithmically effectively addresses fairness gaps to create "locally optimal" models within the original data distribution, this optimality is not true in new test settings. Surprisingly, we find that models with less encoding of demographic attributes are often most "globally optimal", exhibiting better fairness during model evaluation in new test environments. Our work establishes best practices for medical imaging models which maintain their performance and fairness in deployments beyond their initial training contexts, underscoring critical considerations for AI clinical deployments across populations and sites., Comment: Code and data are available at https://github.com/YyzHarry/shortcut-ood-fairness

The outbreak of COVID-19 exposed the inadequacy of our technical tools for home health surveillance, and recent studies have shown the potential of smartphones as a universal optical microscopic imaging platform for such applications. However, most of them use laboratory-grade optomechanical components and transmitted illuminations to ensure focus tuning capability and imaging quality, which keeps the cost of the equipment high. Here we propose an ultra-low-cost solution for smartphone microscopy. To realize focus tunability, we designed a seesaw-like structure capable of converting large displacements on one side into small displacements on the other (reduced to ~9.1%), which leverages the intrinsic flexibility of 3D printing materials. We achieved a focus-tuning accuracy of ~5 micron, which is 40 times higher than the machining accuracy of the 3D-printed lens holder itself. For microscopic imaging, we use an off-the-shelf smartphone camera lens as the objective and the built-in flashlight as the illumination. To compensate for the resulting image quality degradation, we developed a learning-based image enhancement method. We use the CycleGAN architecture to establish the mapping from smartphone microscope images to benchtop microscope images without pairing. We verified the imaging performance on different biomedical samples. Except for the smartphone, we kept the full costs of the device under 4 USD. We think these efforts to lower the costs of smartphone microscopes will benefit their applications in various scenarios, such as point-of-care testing, on-site diagnosis, and home health surveillance.

The recent boom of linear forecasting models questions the ongoing passion for architectural modifications of Transformer-based forecasters. These forecasters leverage Transformers to model the global dependencies over temporal tokens of time series, with each token formed by multiple variates of the same timestamp. However, Transformers are challenged in forecasting series with larger lookback windows due to performance degradation and computation explosion. Besides, the embedding for each temporal token fuses multiple variates that represent potential delayed events and distinct physical measurements, which may fail in learning variate-centric representations and result in meaningless attention maps. In this work, we reflect on the competent duties of Transformer components and repurpose the Transformer architecture without any modification to the basic components. We propose iTransformer that simply applies the attention and feed-forward network on the inverted dimensions. Specifically, the time points of individual series are embedded into variate tokens which are utilized by the attention mechanism to capture multivariate correlations; meanwhile, the feed-forward network is applied for each variate token to learn nonlinear representations. The iTransformer model achieves state-of-the-art on challenging real-world datasets, which further empowers the Transformer family with promoted performance, generalization ability across different variates, and better utilization of arbitrary lookback windows, making it a nice alternative as the fundamental backbone of time series forecasting. Code is available at this repository: https://github.com/thuml/iTransformer.

Machine learning (ML) components are increasingly incorporated into software products, yet developers face challenges in transitioning from ML prototypes to products. Academic researchers struggle to propose solutions to these challenges and evaluate interventions because they often do not have access to close-sourced ML products from industry. In this study, we define and identify open-source ML products, curating a dataset of 262 repositories from GitHub, to facilitate further research and education. As a start, we explore six broad research questions related to different development activities and report 21 findings from a sample of 30 ML products from the dataset. Our findings reveal a variety of development practices and architectural decisions surrounding different types and uses of ML models that offer ample opportunities for future research innovations. We also find very little evidence of industry best practices such as model testing and pipeline automation within the open-source ML products, which leaves room for further investigation to understand its potential impact on the development and eventual end-user experience for the products.

Prevalent in many real-world settings such as healthcare, irregular time series are challenging to formulate predictions from. It is difficult to infer the value of a feature at any given time when observations are sporadic, as it could take on a range of values depending on when it was last observed. To characterize this uncertainty we present EDICT, a strategy that learns an evidential distribution over irregular time series in continuous time. This distribution enables well-calibrated and flexible inference of partially observed features at any time of interest, while expanding uncertainty temporally for sparse, irregular observations. We demonstrate that EDICT attains competitive performance on challenging time series classification tasks and enabling uncertainty-guided inference when encountering noisy data., Comment: ICML 2023 Workshop on Interpretable Machine Learning in Healthcare. Code is available at https://github.com/twkillian/EDICT

The development of remote sensing and deep learning techniques has enabled building semantic segmentation with high accuracy and efficiency. Despite their success in different tasks, the discussions on the impact of spatial resolution on deep learning based building semantic segmentation are quite inadequate, which makes choosing a higher cost-effective data source a big challenge. To address the issue mentioned above, in this study, we create remote sensing images among three study areas into multiple spatial resolutions by super-resolution and down-sampling. After that, two representative deep learning architectures: UNet and FPN, are selected for model training and testing. The experimental results obtained from three cities with two deep learning models indicate that the spatial resolution greatly influences building segmentation results, and with a better cost-effectiveness around 0.3m, which we believe will be an important insight for data selection and preparation.

In recent years, single-frame image super-resolution (SR) has become more realistic by considering the zooming effect and using real-world short- and long-focus image pairs. In this paper, we further investigate the feasibility of applying realistic multi-frame clips to enhance zoom quality via spatio-temporal information coupling. Specifically, we first built a real-world video benchmark, VideoRAW, by a synchronized co-axis optical system. The dataset contains paired short-focus raw and long-focus sRGB videos of different dynamic scenes. Based on VideoRAW, we then presented a Spatio-Temporal Coupling Loss, termed as STCL. The proposed STCL is intended for better utilization of information from paired and adjacent frames to align and fuse features both temporally and spatially at the feature level. The outperformed experimental results obtained in different zoom scenarios demonstrate the superiority of integrating real-world video dataset and STCL into existing SR models for zoom quality enhancement, and reveal that the proposed method can serve as an advanced and viable tool for video zoom., Comment: 11 pages

Out-of-Distribution (OOD) detection is critical for the reliable operation of open-world intelligent systems. Despite the emergence of an increasing number of OOD detection methods, the evaluation inconsistencies present challenges for tracking the progress in this field. OpenOOD v1 initiated the unification of the OOD detection evaluation but faced limitations in scalability and usability. In response, this paper presents OpenOOD v1.5, a significant improvement from its predecessor that ensures accurate, standardized, and user-friendly evaluation of OOD detection methodologies. Notably, OpenOOD v1.5 extends its evaluation capabilities to large-scale datasets such as ImageNet, investigates full-spectrum OOD detection which is important yet underexplored, and introduces new features including an online leaderboard and an easy-to-use evaluator. This work also contributes in-depth analysis and insights derived from comprehensive experimental results, thereby enriching the knowledge pool of OOD detection methodologies. With these enhancements, OpenOOD v1.5 aims to drive advancements and offer a more robust and comprehensive evaluation benchmark for OOD detection research., Comment: Submitted to NeurIPS Datasets & Benchmarks Track. See code at https://github.com/Jingkang50/OpenOOD/ and leaderboard at https://zjysteven.github.io/OpenOOD/

Nonuniform rotational distortion (NURD) correction is vital for endoscopic optical coherence tomography (OCT) imaging and its functional extensions, such as angiography and elastography. Current NURD correction methods require time-consuming feature tracking or cross-correlation calculations and thus sacrifice temporal resolution. Here we propose a cross-attention learning method for the NURD correction in OCT. Our method is inspired by the recent success of the self-attention mechanism in natural language processing and computer vision. By leveraging its ability to model long-range dependencies, we can directly obtain the correlation between OCT A-lines at any distance, thus accelerating the NURD correction. We develop an end-to-end stacked cross-attention network and design three types of optimization constraints. We compare our method with two traditional feature-based methods and a CNN-based method, on two publicly-available endoscopic OCT datasets and a private dataset collected on our home-built endoscopic OCT system. Our method achieved a $\sim3\times$ speedup to real time ($26\pm 3$ fps), and superior correction performance.

Deep learning has been successfully applied to OCT segmentation. However, for data from different manufacturers and imaging protocols, and for different regions of interest (ROIs), it requires laborious and time-consuming data annotation and training, which is undesirable in many scenarios, such as surgical navigation and multi-center clinical trials. Here we propose an annotation-efficient learning method for OCT segmentation that could significantly reduce annotation costs. Leveraging self-supervised generative learning, we train a Transformer-based model to learn the OCT imagery. Then we connect the trained Transformer-based encoder to a CNN-based decoder, to learn the dense pixel-wise prediction in OCT segmentation. These training phases use open-access data and thus incur no annotation costs, and the pre-trained model can be adapted to different data and ROIs without re-training. Based on the greedy approximation for the k-center problem, we also introduce an algorithm for the selective annotation of the target data. We verified our method on publicly-available and private OCT datasets. Compared to the widely-used U-Net model with 100% training data, our method only requires ~10% of the data for achieving the same segmentation accuracy, and it speeds the training up to ~3.5 times. Furthermore, our proposed method outperforms other potential strategies that could improve annotation efficiency. We think this emphasis on learning efficiency may help improve the intelligence and application penetration of OCT-based technologies. Our code and pre-trained model are publicly available at https://github.com/SJTU-Intelligent-Optics-Lab/Annotation-efficient-learning-for-OCT-segmentation., Comment: Submitted to Biomedical Optics Express on January 21, 2023

Incorporating machine learning (ML) components into software products raises new software-engineering challenges and exacerbates existing challenges. Many researchers have invested significant effort in understanding the challenges of industry practitioners working on building products with ML components, through interviews and surveys with practitioners. With the intention to aggregate and present their collective findings, we conduct a meta-summary study: We collect 50 relevant papers that together interacted with over 4758 practitioners using guidelines for systematic literature reviews. We then collected, grouped, and organized the over 500 mentions of challenges within those papers. We highlight the most commonly reported challenges and hope this meta-summary will be a useful resource for the research community to prioritize research and education in this field., Comment: 15 pages, 2 figures, 3 tables, published in CAIN 2023

The chain graph model admits both undirected and directed edges in one graph, where symmetric conditional dependencies are encoded via undirected edges and asymmetric causal relations are encoded via directed edges. Though frequently encountered in practice, the chain graph model has been largely under investigated in literature, possibly due to the lack of identifiability conditions between undirected and directed edges. In this paper, we first establish a set of novel identifiability conditions for the Gaussian chain graph model, exploiting a low rank plus sparse decomposition of the precision matrix. Further, an efficient learning algorithm is built upon the identifiability conditions to fully recover the chain graph structure. Theoretical analysis on the proposed method is conducted, assuring its asymptotic consistency in recovering the exact chain graph structure. The advantage of the proposed method is also supported by numerical experiments on both simulated examples and a real application on the Standard & Poor 500 index data.

Machine learning models often perform poorly on subgroups that are underrepresented in the training data. Yet, little is understood on the variation in mechanisms that cause subpopulation shifts, and how algorithms generalize across such diverse shifts at scale. In this work, we provide a fine-grained analysis of subpopulation shift. We first propose a unified framework that dissects and explains common shifts in subgroups. We then establish a comprehensive benchmark of 20 state-of-the-art algorithms evaluated on 12 real-world datasets in vision, language, and healthcare domains. With results obtained from training over 10,000 models, we reveal intriguing observations for future progress in this space. First, existing algorithms only improve subgroup robustness over certain types of shifts but not others. Moreover, while current algorithms rely on group-annotated validation data for model selection, we find that a simple selection criterion based on worst-class accuracy is surprisingly effective even without any group information. Finally, unlike existing works that solely aim to improve worst-group accuracy (WGA), we demonstrate the fundamental tradeoff between WGA and other important metrics, highlighting the need to carefully choose testing metrics. Code and data are available at: https://github.com/YyzHarry/SubpopBench., Comment: ICML 2023

Time series analysis is widely used in extensive areas. Recently, to reduce labeling expenses and benefit various tasks, self-supervised pre-training has attracted immense interest. One mainstream paradigm is masked modeling, which successfully pre-trains deep models by learning to reconstruct the masked content based on the unmasked part. However, since the semantic information of time series is mainly contained in temporal variations, the standard way of randomly masking a portion of time points will seriously ruin vital temporal variations of time series, making the reconstruction task too difficult to guide representation learning. We thus present SimMTM, a Simple pre-training framework for Masked Time-series Modeling. By relating masked modeling to manifold learning, SimMTM proposes to recover masked time points by the weighted aggregation of multiple neighbors outside the manifold, which eases the reconstruction task by assembling ruined but complementary temporal variations from multiple masked series. SimMTM further learns to uncover the local structure of the manifold, which is helpful for masked modeling. Experimentally, SimMTM achieves state-of-the-art fine-tuning performance compared to the most advanced time series pre-training methods in two canonical time series analysis tasks: forecasting and classification, covering both in- and cross-domain settings.

Longitudinal network consists of a sequence of temporal edges among multiple nodes, where the temporal edges are observed in real time. It has become ubiquitous with the rise of online social platform and e-commerce, but largely under-investigated in literature. In this paper, we propose an efficient estimation framework for longitudinal network, leveraging strengths of adaptive network merging, tensor decomposition and point process. It merges neighboring sparse networks so as to enlarge the number of observed edges and reduce estimation variance, whereas the estimation bias introduced by network merging is controlled by exploiting local temporal structures for adaptive network neighborhood. A projected gradient descent algorithm is proposed to facilitate estimation, where the upper bound of the estimation error in each iteration is established. A thorough analysis is conducted to quantify the asymptotic behavior of the proposed method, which shows that it can significantly reduce the estimation error and also provides guideline for network merging under various scenarios. We further demonstrate the advantage of the proposed method through extensive numerical experiments on synthetic datasets and a militarized interstate dispute dataset., Comment: 30 pages and 4 figures; appendix including technical proof will be uploaded later

Machine learning models frequently experience performance drops under distribution shifts. The underlying cause of such shifts may be multiple simultaneous factors such as changes in data quality, differences in specific covariate distributions, or changes in the relationship between label and features. When a model does fail during deployment, attributing performance change to these factors is critical for the model developer to identify the root cause and take mitigating actions. In this work, we introduce the problem of attributing performance differences between environments to distribution shifts in the underlying data generating mechanisms. We formulate the problem as a cooperative game where the players are distributions. We define the value of a set of distributions to be the change in model performance when only this set of distributions has changed between environments, and derive an importance weighting method for computing the value of an arbitrary set of distributions. The contribution of each distribution to the total performance change is then quantified as its Shapley value. We demonstrate the correctness and utility of our method on synthetic, semi-synthetic, and real-world case studies, showing its effectiveness in attributing performance changes to a wide range of distribution shifts., Comment: Published in ICML 2023

This paper proposes a novel signed $\beta$-model for directed signed network, which is frequently encountered in application domains but largely neglected in literature. The proposed signed $\beta$-model decomposes a directed signed network as the difference of two unsigned networks and embeds each node with two latent factors for in-status and out-status. The presence of negative edges leads to a non-concave log-likelihood, and a one-step estimation algorithm is developed to facilitate parameter estimation, which is efficient both theoretically and computationally. We also develop an inferential procedure for pairwise and multiple node comparisons under the signed $\beta$-model, which fills the void of lacking uncertainty quantification for node ranking. Theoretical results are established for the coverage probability of confidence interval, as well as the false discovery rate (FDR) control for multiple node comparison. The finite sample performance of the signed $\beta$-model is also examined through extensive numerical experiments on both synthetic and real-life networks.

Signed networks are frequently observed in real life with additional sign information associated with each edge, yet such information has been largely ignored in existing network models. This paper develops a unified embedding model for signed networks to disentangle the intertwined balance structure and anomaly effect, which can greatly facilitate the downstream analysis, including community detection, anomaly detection, and network inference. The proposed model captures both balance structure and anomaly effect through a low rank plus sparse matrix decomposition, which are jointly estimated via a regularized formulation. Its theoretical guarantees are established in terms of asymptotic consistency and finite-sample probability bounds for network embedding, community detection and anomaly detection. The advantage of the proposed embedding model is also demonstrated through extensive numerical experiments on both synthetic networks and an international relation network., Comment: 24 pages, 4 figures. The appendix containing technical proof is not included, and will be uploaded in the future

Machine learning models in safety-critical settings like healthcare are often blackboxes: they contain a large number of parameters which are not transparent to users. Post-hoc explainability methods where a simple, human-interpretable model imitates the behavior of these blackbox models are often proposed to help users trust model predictions. In this work, we audit the quality of such explanations for different protected subgroups using real data from four settings in finance, healthcare, college admissions, and the US justice system. Across two different blackbox model architectures and four popular explainability methods, we find that the approximation quality of explanation models, also known as the fidelity, differs significantly between subgroups. We also demonstrate that pairing explainability methods with recent advances in robust machine learning can improve explanation fairness in some settings. However, we highlight the importance of communicating details of non-zero fidelity gaps to users, since a single solution might not exist across all settings. Finally, we discuss the implications of unfair explanation models as a challenging and understudied problem facing the machine learning community., Comment: Published in FAccT 2022

Recently, end-to-end speaker extraction has attracted increasing attention and shown promising results. However, its performance is often inferior to that of a blind source separation (BSS) counterpart with a similar network architecture, due to the auxiliary speaker encoder may sometimes generate ambiguous speaker embeddings. Such ambiguous guidance information may confuse the separation network and hence lead to wrong extraction results, which deteriorates the overall performance. We refer to this as the target confusion problem. In this paper, we conduct an analysis of such an issue and solve it in two stages. In the training phase, we propose to integrate metric learning methods to improve the distinguishability of embeddings produced by the speaker encoder. While for inference, a novel post-filtering strategy is designed to revise the wrong results. Specifically, we first identify these confusion samples by measuring the similarities between output estimates and enrollment utterances, after which the true target sources are recovered by a subtraction operation. Experiments show that performance improvement of more than 1dB SI-SDRi can be brought, which validates the effectiveness of our methods and emphasizes the impact of the target confusion problem., Comment: 5 pages, 1 table, 5 figures. Submitted to INTERSPEECH 2022