Search

Quantile regression and conditional density estimation can reveal structure that is missed by mean regression, such as multimodality and skewness. In this paper, we introduce a deep learning generative model for joint quantile estimation called Penalized Generative Quantile Regression (PGQR). Our approach simultaneously generates samples from many random quantile levels, allowing us to infer the conditional distribution of a response variable given a set of covariates. Our method employs a novel variability penalty to avoid the problem of vanishing variability, or memorization, in deep generative models. Further, we introduce a new family of partial monotonic neural networks (PMNN) to circumvent the problem of crossing quantile curves. A major benefit of PGQR is that it can be fit using a single optimization, thus bypassing the need to repeatedly train the model at multiple quantile levels or use computationally expensive cross-validation to tune the penalty parameter. We illustrate the efficacy of PGQR through extensive simulation studies and analysis of real datasets. Code to implement our method is available at https://github.com/shijiew97/PGQR., Comment: 41 pages, 17 figures, 4 tables. New version includes more simulation studies, comparisons to competing methods, illustrations, real data applications, and discussion of the vanishing variability phenomenon and overparameterization in deep learning. The figures are higher-resolution, and the presentation and writing have improved

Roundoff error problems have occurred frequently in interpolation methods of time-fractional equations, which can lead to undesirable results such as the failure of optimal convergence. These problems are essentially caused by catastrophic cancellations. Currently, a feasible way to avoid these cancellations is using the Gauss--Kronrod quadrature to approximate the integral formulas of coefficients rather than computing the explicit formulas directly for example in the L2-type methods. This nevertheless increases computational cost and arises additional integration errors. In this work, a new framework to handle catastrophic cancellations is proposed, in particular, in the computation of the coefficients for standard and fast L2-type methods on general nonuniform meshes. We propose a concept of $\delta$-cancellation and then some threshold conditions ensuring that $\delta$-cancellations will not happen. If the threshold conditions are not satisfied, a Taylor-expansion technique is proposed to avoid $\delta$-cancellation. Numerical experiments show that our proposed method performs as accurate as the Gauss--Kronrod quadrature method and meanwhile much more efficient. This enables us to complete long time simulations with hundreds of thousands of time steps in short time.

This paper studies efficient algorithms for dynamic curing policies and the corresponding network design problems to guarantee the fast extinction of epidemic spread in a susceptible-infected-susceptible (SIS) model. We consider a Markov process-based SIS epidemic model. We provide a computationally efficient curing algorithm based on the curing policy proposed by Drakopoulos, Ozdaglar, and Tsitsiklis (2014). Since the corresponding optimization problem is NP-hard, finding optimal policies is intractable for large graphs. We provide approximation guarantees on the curing budget of the proposed dynamic curing algorithm. We also present a curing algorithm fair to demographic groups. When the total infection rate is high, the original curing policy includes a waiting period in which no measure is taken to mitigate the spread until the rate slows down. To avoid the waiting period, we study network design problems to reduce the total infection rate by deleting edges or reducing the weight of edges. Then the curing processes become continuous since the total infection rate is restricted by network design. We provide algorithms with provable guarantees for the considered network design problems. In summary, the proposed curing and network design algorithms together provide an effective and computationally efficient approach that mitigates SIS epidemic spread in networks., Comment: 24 pages, 3 figure

Fine-grained object retrieval aims to learn discriminative representation to retrieve visually similar objects. However, existing top-performing works usually impose pairwise similarities on the semantic embedding spaces or design a localization sub-network to continually fine-tune the entire model in limited data scenarios, thus resulting in convergence to suboptimal solutions. In this paper, we develop Fine-grained Retrieval Prompt Tuning (FRPT), which steers a frozen pre-trained model to perform the fine-grained retrieval task from the perspectives of sample prompting and feature adaptation. Specifically, FRPT only needs to learn fewer parameters in the prompt and adaptation instead of fine-tuning the entire model, thus solving the issue of convergence to suboptimal solutions caused by fine-tuning the entire model. Technically, a discriminative perturbation prompt (DPP) is introduced and deemed as a sample prompting process, which amplifies and even exaggerates some discriminative elements contributing to category prediction via a content-aware inhomogeneous sampling operation. In this way, DPP can make the fine-grained retrieval task aided by the perturbation prompts close to the solved task during the original pre-training. Thereby, it preserves the generalization and discrimination of representation extracted from input samples. Besides, a category-specific awareness head is proposed and regarded as feature adaptation, which removes the species discrepancies in features extracted by the pre-trained model using category-guided instance normalization. And thus, it makes the optimized features only include the discrepancies among subcategories. Extensive experiments demonstrate that our FRPT with fewer learnable parameters achieves the state-of-the-art performance on three widely-used fine-grained datasets., Comment: Accepted by AAAI 2023

Recent ALMA observations on disk substructures suggest the presence of embedded protoplanets in a large number disks. The primordial configurations of these planetary systems can be deduced from the morphology of the disk substructure and serve as initial conditions for numerical investigation of their future evolution. Starting from the initial configurations of 12 multi-planetary systems deduced from ALMA disks, we carried out two-stage N-body simulation to investigate the evolution of the planetary systems at the disk stage as well as the long term orbital stability after the disk dispersal. At the disk stage, our simulation includes both the orbital migration and pebble/gas accretion effects. We found a variety of planetary systems are produced and can be categorised into distant giant planets, Jupiter-like planets, Neptune-like planets and distant small planets. We found the disk stage evolution as well as the final configurations are sensitive to both the initial mass assignments and viscosity. After the disk stage, we implement only mutual gravity between star and planets and introduce stochastic perturbative forces. All systems are integrated for up to 10 Gyr to test their orbital stability. Most planetary systems are found to be stable for at least 10 Gyr with perturbative force in a reasonable range. Our result implies that a strong perturbation source such as stellar flybys is required to drive the planetary system unstable. We discuss the implications of our results on both the disk and planet observation, which may be confirmed by the next generation telescopes such as JWST and ngVLA., Comment: 41 pages, 16 figures, accepted for publication in ApJ

We present an approach to efficiently and effectively adapt a masked image modeling (MIM) pre-trained vanilla Vision Transformer (ViT) for object detection, which is based on our two novel observations: (i) A MIM pre-trained vanilla ViT encoder can work surprisingly well in the challenging object-level recognition scenario even with randomly sampled partial observations, e.g., only 25% $\sim$ 50% of the input embeddings. (ii) In order to construct multi-scale representations for object detection from single-scale ViT, a randomly initialized compact convolutional stem supplants the pre-trained large kernel patchify stem, and its intermediate features can naturally serve as the higher resolution inputs of a feature pyramid network without further upsampling or other manipulations. While the pre-trained ViT is only regarded as the 3$^{rd}$-stage of our detector's backbone instead of the whole feature extractor. This results in a ConvNet-ViT hybrid feature extractor. The proposed detector, named MIMDet, enables a MIM pre-trained vanilla ViT to outperform hierarchical Swin Transformer by 2.5 box AP and 2.6 mask AP on COCO, and achieves better results compared with the previous best adapted vanilla ViT detector using a more modest fine-tuning recipe while converging 2.8$\times$ faster. Code and pre-trained models are available at https://github.com/hustvl/MIMDet., Comment: v2: more analysis & stronger results. Preprint. Work in progress. Code and pre-trained models are available at https://github.com/hustvl/MIMDet

Abstract: Bionic learning with fused sensing, memory and processing functions outperforms artificial neural networks running on silicon chips in terms of efficiency and footprint. However, digital hardware implementation of bionic learning suffers from device heterogeneity in sensors and processing cores, which incurs large hardware, energy and time overheads. Here, we present a universal solution to simultaneously perform multi-modal sensing, memory and processing using organic electrochemical transistors with designed architecture and tailored channel morphology, selective ion injection into the crystalline/amorphous regions. The resultant device work as either a volatile receptor that shows multi-modal sensing, or a non-volatile synapse that features record-high 10-bit analog states, low switching stochasticity and good retention without the integration of any extra devices. Homogeneous integration of such devices enables bionic learning functions such as conditioned reflex and real-time cardiac disease diagnose via reservoir computing, illustrating the promise for future smart edge health informatics., Comment: 17 pages, 4 figures

The integration of high-k dielectrics with two-dimensional (2D) semiconductors is a critical step towards high-performance nanoelectronics, which however remains challenging due to high density of interface states and the damage to the monolayer 2D semiconductors. In this study, we propose a selective hydrogenation strategy to improve the interface properties while do not affect the 2D semiconductors. Using the interface of monolayer MoS2 and silicon nitride as an example, we show substantially improved interface properties for electronic applications after the interfacial hydrogenation, as evidenced by reduced inhomogeneous charge redistribution, increased band offset, and untouched electronic properties of MoS2. Interestingly, this hydrogenation process selectively occurs only at the silicon nitride surface and is compatible with the current semiconductor fabrication process. We further show that this strategy is general and applicable to other interfaces between high-k dielectrics and 2D semiconductors such as HfO2 on the monolayer MoS2. Our results demonstrate a simple yet viable way to improve the interfacial properties for integrating many high-k dielectrics on a broad range of two-dimensional transition metal disulfide semiconductors.

Recent ALMA observations have identified a variety of dust gaps in protoplanetary disks, which are commonly interpreted to be generated by unobserved planets. Predicting mass of such embedded planets is of fundamental importance in comparing those disk architectures with the observed diversity of exoplanets. The prediction, however, depends on the assumption that whether the same gap structure exists in the dust component alone or in the gas component as well. We assume a planet can only open a gap in the gas component when its mass exceeds the pebble isolation mass by considering the core accretion scenario. We then propose two criteria to distinguish if a gap is opened in the dust disk alone or the gas gap as well when observation data on the gas profile is not available. We apply the criteria to 35 disk systems with a total of 55 gaps compiled from previous studies, and classify each gap into four different groups. The classification of the observed gaps allows us to predict the mass of embedded planets in a consistent manner with the pebble isolation mass. We find that outer gaps are mostly dust alone, while inner gaps are more likely to be associated with a gas gap as well. The distribution of such embedded planets is very different from the architecture of the observed planetary systems, suggesting that the significant inward migration is required in their evolution., Comment: 25 pages, 8 figures, accepted for publication in ApJ; sec 2 revised, figures and tables updated