arXiv subjects

Feng Chen

Publications and source records attributed to Feng Chen.

At least 271 records · Page 15Linked to original sources

arXiv2022

Framing Algorithmic Recourse for Anomaly Detection

The problem of algorithmic recourse has been explored for supervised machine learning models, to provide more interpretable, transparent and robust outcomes from decision support systems. An unexplored area is that of algorithmic recourse for anomaly detection, specifically for tabular data with only discrete feature values. Here the problem is to present a set of counterfactuals that are deemed normal by the underlying anomaly detection model so that applications can utilize this information for explanation purposes or to recommend countermeasures. We present an approach -- Context preserving Algorithmic Recourse for Anomalies in Tabular data (CARAT), that is effective, scalable, and agnostic to the underlying anomaly detection model. CARAT uses a transformer based encoder-decoder model to explain an anomaly by finding features with low likelihood. Subsequently semantically coherent counterfactuals are generated by modifying the highlighted features, using the overall context of features in the anomalous instance(s). Extensive experiments help demonstrate the efficacy of CARAT.

cs.LG↗

arXiv2022

Radiative Magnetohydrodynamic Simulation of the Confined Eruption of a Magnetic Flux Rope: Magnetic Structure and Plasma Thermodynamics

It is widely believed that magnetic flux ropes are the key structure of solar eruptions; however, their observable counterparts are not clear yet. We study a flare associated with flux rope eruption in a comprehensive radiative magnetohydrodynamic simulation of flare-productive active regions, especially focusing on the thermodynamic properties of the plasma involved in the eruption and their relation to the magnetic flux rope. The pre-existing flux rope, which carries cold and dense plasma, rises quasi-statically before the eruption onsets. During this stage, the flux rope does not show obvious signatures in extreme ultraviolet (EUV) emission. After the flare onset, a thin `current shell' is generated around the erupting flux rope. Moreover, a current sheet is formed under the flux rope, where two groups of magnetic arcades reconnect and create a group of post-flare loops. The plasma within the `current shell', current sheet, and post-flare loops are heated to more than 10 MK. The post-flare loops give rise to abundant soft X-ray emission. Meanwhile a majority of the plasma hosted in the flux rope is heated to around 1 MK, and the main body of the flux rope is manifested as a bright arch in cooler EUV passbands such as AIA 171 Å~channel.

astro-ph.SR↗

arXiv2022

Calibrated Nonparametric Scan Statistics for Anomalous Pattern Detection in Graphs

We propose a new approach, the calibrated nonparametric scan statistic (CNSS), for more accurate detection of anomalous patterns in large-scale, real-world graphs. Scan statistics identify connected subgraphs that are interesting or unexpected through maximization of a likelihood ratio statistic; in particular, nonparametric scan statistics (NPSSs) identify subgraphs with a higher than expected proportion of individually significant nodes. However, we show that recently proposed NPSS methods are miscalibrated, failing to account for the maximization of the statistic over the multiplicity of subgraphs. This results in both reduced detection power for subtle signals, and low precision of the detected subgraph even for stronger signals. Thus we develop a new statistical approach to recalibrate NPSSs, correctly adjusting for multiple hypothesis testing and taking the underlying graph structure into account. While the recalibration, based on randomization testing, is computationally expensive, we propose both an efficient (approximate) algorithm and new, closed-form lower bounds (on the expected maximum proportion of significant nodes for subgraphs of a given size, under the null hypothesis of no anomalous patterns). These advances, along with the integration of recent core-tree decomposition methods, enable CNSS to scale to large real-world graphs, with substantial improvement in the accuracy of detected subgraphs. Extensive experiments on both semi-synthetic and real-world datasets are demonstrated to validate the effectiveness of our proposed methods, in comparison with state-of-the-art counterparts.

stat.ME↗

arXiv2022

A Survey on Uncertainty Reasoning and Quantification for Decision Making: Belief Theory Meets Deep Learning

An in-depth understanding of uncertainty is the first step to making effective decisions under uncertainty. Deep/machine learning (ML/DL) has been hugely leveraged to solve complex problems involved with processing high-dimensional data. However, reasoning and quantifying different types of uncertainties to achieve effective decision-making have been much less explored in ML/DL than in other Artificial Intelligence (AI) domains. In particular, belief/evidence theories have been studied in KRR since the 1960s to reason and measure uncertainties to enhance decision-making effectiveness. We found that only a few studies have leveraged the mature uncertainty research in belief/evidence theories in ML/DL to tackle complex problems under different types of uncertainty. In this survey paper, we discuss several popular belief theories and their core ideas dealing with uncertainty causes and types and quantifying them, along with the discussions of their applicability in ML/DL. In addition, we discuss three main approaches that leverage belief theories in Deep Neural Networks (DNNs), including Evidential DNNs, Fuzzy DNNs, and Rough DNNs, in terms of their uncertainty causes, types, and quantification methods along with their applicability in diverse problem domains. Based on our in-depth survey, we discuss insights, lessons learned, limitations of the current state-of-the-art bridging belief theories and ML/DL, and finally, future research directions.

cs.AI↗

arXiv2022

Parallel plasma loops and the energization of the solar corona

The outer atmosphere of the Sun is composed of plasma heated to temperatures well in excess of the visible surface. We investigate short cool and warm (<1 MK) loops seen in the core of an active region to address the role of field-line braiding in energising these structures. We report observations from the High-resolution Coronal imager (Hi-C) that have been acquired in a coordinated campaign with the Interface Region Imaging Spectrograph (IRIS). In the core of the active region, the 172 A band of Hi-C and the 1400 A channel of IRIS show plasma loops at different temperatures that run in parallel. There is a small but detectable spatial offset of less than 1 arcsec between the loops seen in the two bands. Most importantly, we do not see observational signatures that these loops might be twisted around each other. Considering the scenario of magnetic braiding, our observations of parallel loops imply that the stresses put into the magnetic field have to relax while the braiding is applied: the magnetic field never reaches a highly braided state on these length-scales comparable to the separation of the loops. This supports recent numerical 3D models of loop braiding in which the effective dissipation is sufficiently large that it keeps the magnetic field from getting highly twisted within a loop.

astro-ph.SR↗

arXiv2022

Adaptive Fairness-Aware Online Meta-Learning for Changing Environments

The fairness-aware online learning framework has arisen as a powerful tool for the continual lifelong learning setting. The goal for the learner is to sequentially learn new tasks where they come one after another over time and the learner ensures the statistic parity of the new coming task across different protected sub-populations (e.g. race and gender). A major drawback of existing methods is that they make heavy use of the i.i.d assumption for data and hence provide static regret analysis for the framework. However, low static regret cannot imply a good performance in changing environments where tasks are sampled from heterogeneous distributions. To address the fairness-aware online learning problem in changing environments, in this paper, we first construct a novel regret metric FairSAR by adding long-term fairness constraints onto a strongly adapted loss regret. Furthermore, to determine a good model parameter at each round, we propose a novel adaptive fairness-aware online meta-learning algorithm, namely FairSAOML, which is able to adapt to changing environments in both bias control and model precision. The problem is formulated in the form of a bi-level convex-concave optimization with respect to the model's primal and dual parameters that are associated with the model's accuracy and fairness, respectively. The theoretic analysis provides sub-linear upper bounds for both loss regret and violation of cumulative fairness constraints. Our experimental evaluation on different real-world datasets with settings of changing environments suggests that the proposed FairSAOML significantly outperforms alternatives based on the best prior online learning approaches.

cs.LG↗

arXiv2022

S4OD: Semi-Supervised learning for Single-Stage Object Detection

Single-stage detectors suffer from extreme foreground-background class imbalance, while two-stage detectors do not. Therefore, in semi-supervised object detection, two-stage detectors can deliver remarkable performance by only selecting high-quality pseudo labels based on classification scores. However, directly applying this strategy to single-stage detectors would aggravate the class imbalance with fewer positive samples. Thus, single-stage detectors have to consider both quality and quantity of pseudo labels simultaneously. In this paper, we design a dynamic self-adaptive threshold (DSAT) strategy in classification branch, which can automatically select pseudo labels to achieve an optimal trade-off between quality and quantity. Besides, to assess the regression quality of pseudo labels in single-stage detectors, we propose a module to compute the regression uncertainty of boxes based on Non-Maximum Suppression. By leveraging only 10% labeled data from COCO, our method achieves 35.0% AP on anchor-free detector (FCOS) and 32.9% on anchor-based detector (RetinaNet).

cs.CV↗

arXiv2022

Lane detection with Position Embedding

Recently, lane detection has made great progress in autonomous driving. RESA (REcurrent Feature-Shift Aggregator) is based on image segmentation. It presents a novel module to enrich lane feature after preliminary feature extraction with an ordinary CNN. For Tusimple dataset, there is not too complicated scene and lane has more prominent spatial features. On the basis of RESA, we introduce the method of position embedding to enhance the spatial features. The experimental results show that this method has achieved the best accuracy 96.93% on Tusimple dataset.

cs.CV↗

arXiv2022

Interacting Attention Graph for Single Image Two-Hand Reconstruction

Graph convolutional network (GCN) has achieved great success in single hand reconstruction task, while interacting two-hand reconstruction by GCN remains unexplored. In this paper, we present Interacting Attention Graph Hand (IntagHand), the first graph convolution based network that reconstructs two interacting hands from a single RGB image. To solve occlusion and interaction challenges of two-hand reconstruction, we introduce two novel attention based modules in each upsampling step of the original GCN. The first module is the pyramid image feature attention (PIFA) module, which utilizes multiresolution features to implicitly obtain vertex-to-image alignment. The second module is the cross hand attention (CHA) module that encodes the coherence of interacting hands by building dense cross-attention between two hand vertices. As a result, our model outperforms all existing two-hand reconstruction methods by a large margin on InterHand2.6M benchmark. Moreover, ablation studies verify the effectiveness of both PIFA and CHA modules for improving the reconstruction accuracy. Results on in-the-wild images and live video streams further demonstrate the generalization ability of our network. Our code is available at https://github.com/Dw1010/IntagHand.

cs.CV↗

arXiv2022

Multi-Agent Policy Transfer via Task Relationship Modeling

Team adaptation to new cooperative tasks is a hallmark of human intelligence, which has yet to be fully realized in learning agents. Previous work on multi-agent transfer learning accommodate teams of different sizes, heavily relying on the generalization ability of neural networks for adapting to unseen tasks. We believe that the relationship among tasks provides the key information for policy adaptation. In this paper, we try to discover and exploit common structures among tasks for more efficient transfer, and propose to learn effect-based task representations as a common space of tasks, using an alternatively fixed training scheme. We demonstrate that the task representation can capture the relationship among tasks, and can generalize to unseen tasks. As a result, the proposed method can help transfer learned cooperation knowledge to new tasks after training on a few source tasks. We also find that fine-tuning the transferred policies help solve tasks that are hard to learn from scratch.

cs.AI↗

arXiv2022

Layer Adaptive Deep Neural Networks for Out-of-distribution Detection

During the forward pass of Deep Neural Networks (DNNs), inputs gradually transformed from low-level features to high-level conceptual labels. While features at different layers could summarize the important factors of the inputs at varying levels, modern out-of-distribution (OOD) detection methods mostly focus on utilizing their ending layer features. In this paper, we proposed a novel layer-adaptive OOD detection framework (LA-OOD) for DNNs that can fully utilize the intermediate layers' outputs. Specifically, instead of training a unified OOD detector at a fixed ending layer, we train multiple One-Class SVM OOD detectors simultaneously at the intermediate layers to exploit the full spectrum characteristics encoded at varying depths of DNNs. We develop a simple yet effective layer-adaptive policy to identify the best layer for detecting each potential OOD example. LA-OOD can be applied to any existing DNNs and does not require access to OOD samples during the training. Using three DNNs of varying depth and architectures, our experiments demonstrate that LA-OOD is robust against OODs of varying complexity and can outperform state-of-the-art competitors by a large margin on some real-world datasets.

cs.LG↗

arXiv2022

Transfering Hierarchical Structure with Dual Meta Imitation Learning

Hierarchical Imitation Learning (HIL) is an effective way for robots to learn sub-skills from long-horizon unsegmented demonstrations. However, the learned hierarchical structure lacks the mechanism to transfer across multi-tasks or to new tasks, which makes them have to learn from scratch when facing a new situation. Transferring and reorganizing modular sub-skills require fast adaptation ability of the whole hierarchical structure. In this work, we propose Dual Meta Imitation Learning (DMIL), a hierarchical meta imitation learning method where the high-level network and sub-skills are iteratively meta-learned with model-agnostic meta-learning. DMIL uses the likelihood of state-action pairs from each sub-skill as the supervision for the high-level network adaptation, and use the adapted high-level network to determine different data set for each sub-skill adaptation. We theoretically prove the convergence of the iterative training process of DMIL and establish the connection between DMIL and Expectation-Maximization algorithm. Empirically, we achieve state-of-the-art few-shot imitation learning performance on the Meta-world \cite{metaworld} benchmark and competitive results on long-horizon tasks of Kitchen environments.

cs.RO↗

arXiv2022

SEED: Sound Event Early Detection via Evidential Uncertainty

Sound Event Early Detection (SEED) is an essential task in recognizing the acoustic environments and soundscapes. However, most of the existing methods focus on the offline sound event detection, which suffers from the over-confidence issue of early-stage event detection and usually yield unreliable results. To solve the problem, we propose a novel Polyphonic Evidential Neural Network (PENet) to model the evidential uncertainty of the class probability with Beta distribution. Specifically, we use a Beta distribution to model the distribution of class probabilities, and the evidential uncertainty enriches uncertainty representation with evidence information, which plays a central role in reliable prediction. To further improve the event detection performance, we design the backtrack inference method that utilizes both the forward and backward audio features of an ongoing event. Experiments on the DESED database show that the proposed method can simultaneously improve 13.0\% and 3.8\% in time delay and detection F1 score compared to the state-of-the-art methods.

cs.SD↗

arXiv2022

Learning Invariable Semantical Representation from Language for Extensible Policy Generalization

Recently, incorporating natural language instructions into reinforcement learning (RL) to learn semantically meaningful representations and foster generalization has caught many concerns. However, the semantical information in language instructions is usually entangled with task-specific state information, which hampers the learning of semantically invariant and reusable representations. In this paper, we propose a method to learn such representations called element randomization, which extracts task-relevant but environment-agnostic semantics from instructions using a set of environments with randomized elements, e.g., topological structures or textures, yet the same language instruction. We theoretically prove the feasibility of learning semantically invariant representations through randomization. In practice, we accordingly develop a hierarchy of policies, where a high-level policy is designed to modulate the behavior of a goal-conditioned low-level policy by proposing subgoals as semantically invariant representations. Experiments on challenging long-horizon tasks show that (1) our low-level policy reliably generalizes to tasks against environment changes; (2) our hierarchical policy exhibits extensible generalization in unseen new tasks that can be decomposed into several solvable sub-tasks; and (3) by storing and replaying language trajectories as succinct policy representations, the agent can complete tasks in a one-shot fashion, i.e., once one successful trajectory has been attained.

cs.CL↗

arXiv2021

Effects of the initial perturbations on the Rayleigh-Taylor-Kelvin-Helmholtz instability system

In the paper, the effects of initial perturbations on the Rayleigh-Taylor instability (RTI), Kelvin-Helmholtz instability (KHI), and the coupled Rayleigh-Taylor-Kelvin-Helmholtz instability (RTKHI) systems are investigated using a multiple-relaxation-time discrete Boltzmann model. Six different perturbation interfaces are designed to study the effects of the initial perturbations on the instability systems. Based on the mean heat flux strength $D_{3,1}$, the effects of initial interfaces on the coupled RTKHI are examined in detail. The research is focused on two aspects: (i) the main mechanism in the early stage of the RTKHI, (ii) the transition point from KHI-like to RTI-like for the case where the KHI dominates at earlier time and the RTI dominates at later time. It is found that the early main mechanism is related to the shape of the initial interface, which is represented by both the bilateral contact angle $θ_{1}$ and the middle contact angle $θ_{2}$. The influence of inverted parabolic and inverted ellipse perturbations ($θ_{1}<90$) on the transition point of the RTKHI system is greater than that of other interfaces.

physics.flu-dyn↗

arXiv2021

Disambiguation of Vector Magnetograms by Stereoscopic Observations from the Solar Orbiter/Polarimetric and Helioseismic Imager (PHI) and the Solar Dynamic Observatory (SDO)/Helioseismic and Magnetic Imager (HMI)

Spectropolarimetric reconstructions of the photospheric vector magnetic field are intrinsically limited by the so-called 180$^\circ$ ambiguity in the orientation of the transverse component. The successful launch and operation of Solar Orbiter has made the removal of the 180$^\circ$ ambiguity possible using solely observations obtained from two different vantage points. While the exploitation of such a possibility is straightforward in principle, it is less so in practice and it is therefore important to assess the accuracy and limitations, as a function of both the satellites orbits and measurement principles. In this work we present a stereoscopic disambiguation method (SDM) and discuss a thorough testing of its accuracy in applications to modeled active regions and quiet Sun observations. In a first series of tests, we employ magnetograms extracted from three different numerical simulations as test fields, and model observations of the magnetograms from different angles and distances. In these more idealized tests, the SDM is proven to to reach a 100% disambiguation accuracy when applied to moderately-to-well resolved fields. Even in the case of disambiguation of quiet Sun magnetograms with significant under-resolved scale, the SDM provides an accuracy between 82% and 98% depending on the field strength. The accuracy of the SDM is found to be mostly sensitive to the variable resolution of Solar Orbiter on its highly elliptic orbit, as well as to the intrinsic scale of the observed field. Finally, as a more realistic test, we consider magnetograms that are obtained using a radiative transfer inversion code and the SOPHISM instrument simulator applied to a 3D simulation of a pore, and present a preliminary discussion of the effect of the viewing angle on the observed field.

astro-ph.SR↗

arXiv2021

Spatiotemporal ETAS model with a renewal main-shock arrival process

This article proposes a spatiotemporal point process model that enhances the classical Epidemic-Type Aftershock Sequence (ETAS) model by incorporating a renewal main-shock arrival process, which we term the renewal ETAS (RETAS) model. This modification is similar in spirit to the renewal Hawkes (RHawkes) process but the conditional intensity process supports a spatial domain. It empowers the main-shock intensity with the capability to reset upon the arrival of main-shocks and therefore allows for heavier clustering of earthquakes than the spatiotemporal ETAS model introduced by Ogata (1998). We introduce a likelihood evaluation algorithm for parameter estimation and provide a novel procedure to evaluated the fitted model's goodness-of-fit based on a sequential application of the Rosenblatt transformation. A simulation algorithm for the RETAS model is developed and applied to validate the numerical performance of the likelihood evaluation algorithm and goodness of fit test procedure. We illustrate the proposed model and procedures on various earthquake catalogs around the world each with distinctly different seismic activity. These catalogs will demonstrate that the RETAS model affords additional flexibility in comparison to the classical spatiotemporal ETAS model and has the potential for superior modeling and forecasting of seismicity.

stat.ME↗

arXiv2021

Prototypical Model with Novel Information-theoretic Loss Function for Generalized Zero Shot Learning

Generalized zero shot learning (GZSL) is still a technical challenge of deep learning as it has to recognize both source and target classes without data from target classes. To preserve the semantic relation between source and target classes when only trained with data from source classes, we address the quantification of the knowledge transfer and semantic relation from an information-theoretic viewpoint. To this end, we follow the prototypical model and format the variables of concern as a probability vector. Leveraging on the proposed probability vector representation, the information measurement such as mutual information and entropy, can be effectively evaluated with simple closed forms. We discuss the choice of common embedding space and distance function when using the prototypical model. Then We propose three information-theoretic loss functions for deterministic GZSL model: a mutual information loss to bridge seen data and target classes; an uncertainty-aware entropy constraint loss to prevent overfitting when using seen data to learn the embedding of target classes; a semantic preserving cross entropy loss to preserve the semantic relation when mapping the semantic representations to the common space. Simulation shows that, as a deterministic model, our proposed method obtains state of the art results on GZSL benchmark datasets. We achieve 21%-64% improvements over the baseline model -- deep calibration network (DCN) and for the first time demonstrate a deterministic model can perform as well as generative ones. Moreover, our proposed model is compatible with generative models. Simulation studies show that by incorporating with f-CLSWGAN, we obtain comparable results compared with advanced generative models.

cs.LG↗