Characterizing and Avoiding Negative Transfer Wang, Zirui; Dai, Zihang; Poczos, Barnabas ...
2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),
06/2019
Conference Proceeding
Odprti dostop
When labeled data is scarce for a specific target task, transfer learning often offers an effective solution by utilizing data from a related source task. However, when transferring knowledge from a ...less related source, it may inversely hurt the target performance, a phenomenon known as negative transfer. Despite its pervasiveness, negative transfer is usually described in an informal manner, lacking rigorous definition, careful analysis, or systematic treatment. This paper proposes a formal definition of negative transfer and analyzes three important aspects thereof. Stemming from this analysis, a novel technique is proposed to circumvent negative transfer by filtering out unrelated source data. Based on adversarial networks, the technique is highly generic and can be applied to a wide range of transfer learning algorithms. The proposed approach is evaluated on six state-of-the-art deep transfer methods via experiments on four benchmark datasets with varying levels of difficulty. Empirically, the proposed method consistently improves the performance of all baseline methods and largely avoids negative transfer, even when the source data is degenerate.
Deep generative models for galaxy image simulations Lanusse, François; Mandelbaum, Rachel; Ravanbakhsh, Siamak ...
Monthly Notices of the Royal Astronomical Society,
07/2021, Letnik:
504, Številka:
4
Journal Article
Recenzirano
Odprti dostop
ABSTRACT
Image simulations are essential tools for preparing and validating the analysis of current and future wide-field optical surveys. However, the galaxy models used as the basis for these ...simulations are typically limited to simple parametric light profiles, or use a fairly limited amount of available space-based data. In this work, we propose a methodology based on deep generative models to create complex models of galaxy morphologies that may meet the image simulation needs of upcoming surveys. We address the technical challenges associated with learning this morphology model from noisy and point spread function (PSF)-convolved images by building a hybrid Deep Learning/physical Bayesian hierarchical model for observed images, explicitly accounting for the PSF and noise properties. The generative model is further made conditional on physical galaxy parameters, to allow for sampling new light profiles from specific galaxy populations. We demonstrate our ability to train and sample from such a model on galaxy postage stamps from the HST/ACS COSMOS survey, and validate the quality of the model using a range of second- and higher order morphology statistics. Using this set of statistics, we demonstrate significantly more realistic morphologies using these deep generative models compared to conventional parametric models. To help make these generative models practical tools for the community, we introduce galsim-hub, a community-driven repository of generative models, and a framework for incorporating generative models within the galsim image simulation software.
In many scientific and engineering applications, we are tasked with the maximisation of an expensive to evaluate black box function f. Traditional settings for this problem assume just the ...availability of this single function. However, in many cases, cheap approximations to f may be obtainable. For example, the expensive real world behaviour of a robot can be approximated by a cheap computer simulation. We can use these approximations to eliminate low function value regions cheaply and use the expensive evaluations of f in a small but promising region and speedily identify the optimum. We formalise this task as a multi-fidelity bandit problem where the target function and its approximations are sampled from a Gaussian process. We develop MF-GP-UCB, a novel method based on upper confidence bound techniques. In our theoretical analysis we demonstrate that it exhibits precisely the above behaviour and achieves better bounds on the regret than strategies which ignore multi-fidelity information. Empirically, MF-GP-UCB outperforms such naive strategies and other multi-fidelity methods on several synthetic and real experiments.
Background: In the human genome, distal enhancers are involved in regulating target genes through proximal promoters by forming enhancer-promoter interactions. Although recently developed ...high-throughput experimental approaches have allowed us to recognize potential enhancer-promoter interactions genome-wide, it is still largely unclear to what extent the sequence-level information encoded in our genome help guide such interactions.
Methods: Here we report a new computational method (named "SPEID") using deep learning models to predict enhancer-promoter interactions based on sequence-based features only, when the locations of putative enhancers and promoters in a particular cell type are given.
Results: Our results across six different cell types demonstrate that SPEID is effective in predicting enhancer-promoter interactions as compared to state-of-the-art methods that only use information from a single cell type. As a proof-of-principle, we also applied SPEID to identify somatic non-coding mutations in melanoma samples that may have reduced enhancer-promoter interactions in tumor genomes.
Conclusions: This work demonstrates that deep learning models can help reveal that sequence-based features alone are sufficient to reliably predict enhancer-promoter interactions genome-wide.
Abstract
Galaxy-scale strong gravitational lensing can not only provide a valuable probe of the dark matter distribution of massive galaxies, but also provide valuable cosmological constraints, ...either by studying the population of strong lenses or by measuring time delays in lensed quasars. Due to the rarity of galaxy-scale strongly lensed systems, fast and reliable automated lens finding methods will be essential in the era of large surveys such as Large Synoptic Survey Telescope, Euclid and Wide-Field Infrared Survey Telescope. To tackle this challenge, we introduce CMU DeepLens, a new fully automated galaxy–galaxy lens finding method based on deep learning. This supervised machine learning approach does not require any tuning after the training step which only requires realistic image simulations of strongly lensed systems. We train and validate our model on a set of 20 000 LSST-like mock observations including a range of lensed systems of various sizes and signal-to-noise ratios (S/N). We find on our simulated data set that for a rejection rate of non-lenses of 99 per cent, a completeness of 90 per cent can be achieved for lenses with Einstein radii larger than 1.4 arcsec and S/N larger than 20 on individual g-band LSST exposures. Finally, we emphasize the importance of realistically complex simulations for training such machine learning methods by demonstrating that the performance of models of significantly different complexities cannot be distinguished on simpler simulations. We make our code publicly available at https://github.com/McWilliamsCenter/CMUDeepLens.
A common problem in disciplines of applied Statistics research such as Astrostatistics is of estimating the posterior distribution of relevant parameters. Typically, the likelihoods for such models ...are computed via expensive experiments such as cosmological simulations of the universe. An urgent challenge in these research domains is to develop methods that can estimate the posterior with few likelihood evaluations.
In this paper, we study active posterior estimation in a Bayesian setting when the likelihood is expensive to evaluate. Existing techniques for posterior estimation are based on generating samples representative of the posterior. Such methods do not consider efficiency in terms of likelihood evaluations. In order to be query efficient we treat posterior estimation in an active regression framework. We propose two myopic query strategies to choose where to evaluate the likelihood and implement them using Gaussian processes. Via experiments on a series of synthetic and real examples we demonstrate that our approach is significantly more query efficient than existing techniques and other heuristics for posterior estimation.
Innovations in batteries can require years of experimentation for design and optimization. We report an autonomous approach to the optimization of a battery electrolyte that uses machine learning ...coupled to a robotic test-stand to perform hundreds of sequential experiments. We search for mixtures of salts in aqueous electrolytes with high electrochemical stability using Bayesian optimization. In 40 hours of experimentation testing for 140 electrolyte formulas, we converge on a non-intuitive optimal electrolyte. The optimum is a mixed-anion sodium electrolyte that is more stable than a benchmark electrolyte, despite lower salt content, contrary to the known design principle. The precision and repeatability of the robotic test-stand distinguishes formulations that human-guided design may have missed. Our result demonstrates the possibility of integrating robotics with machine learning to discover novel battery materials. We provide a dataset characterizing 251 aqueous electrolytes containing LiNO3, LiClO4, Li2SO4, NaNO3, NaClO4, and Na2SO4 that includes conductivities, pHs, and electrochemical responses on platinum.
Display omitted
Robotics and machine learning combine to autonomously find aqueous electrolytesA non-intuitive, mixed-anion sodium electrolyte is discovered251 mixed-salt electrolytes’ conductivities, electrochemical behavior, and pHs are examined
Dave et al. combine robotics and machine learning to discover aqueous electrolytes. The algorithm runs the experiment without human intervention, delivering a mixed-anion sodium electrolyte with non-intuitively better performance compared to the benchmark. A dataset of 251 blended-salt aqueous electrolytes that includes conductivities, pHs, and electrochemistry electrolytes is obtained.
While deep learning methods have achieved state-of-theart performance in many challenging inverse problems like image inpainting and super-resolution, they invariably involve problem-specific ...training of the networks. Under this approach, each inverse problem requires its own dedicated network. In scenarios where we need to solve a wide variety of problems, e.g., on a mobile camera, it is inefficient and expensive to use these problem-specific networks. On the other hand, traditional methods using analytic signal priors can be used to solve any linear inverse problem; this often comes with a performance that is worse than learning-based methods. In this work, we provide a middle ground between the two kinds of methods - we propose a general framework to train a single deep neural network that solves arbitrary linear inverse problems. We achieve this by training a network that acts as a quasi-projection operator for the set of natural images and show that any linear inverse problem involving natural images can be solved using iterative methods. We empirically show that the proposed framework demonstrates superior performance over traditional methods using wavelet sparsity prior while achieving performance comparable to specially-trained networks on tasks including compressive sensing and pixel-wise inpainting.
Quantifying differences or similarities in connectomes has been a challenge due to the immense complexity of global brain networks. Here we introduce a noninvasive method that uses diffusion MRI to ...characterize whole-brain white matter architecture as a single local connectome fingerprint that allows for a direct comparison between structural connectomes. In four independently acquired data sets with repeated scans (total N = 213), we show that the local connectome fingerprint is highly specific to an individual, allowing for an accurate self-versus-others classification that achieved 100% accuracy across 17,398 identification tests. The estimated classification error was approximately one thousand times smaller than fingerprints derived from diffusivity-based measures or region-to-region connectivity patterns for repeat scans acquired within 3 months. The local connectome fingerprint also revealed neuroplasticity within an individual reflected as a decreasing trend in self-similarity across time, whereas this change was not observed in the diffusivity measures. Moreover, the local connectome fingerprint can be used as a phenotypic marker, revealing 12.51% similarity between monozygotic twins, 5.14% between dizygotic twins, and 4.51% between none-twin siblings, relative to differences between unrelated subjects. This novel approach opens a new door for probing the influence of pathological, genetic, social, or environmental factors on the unique configuration of the human connectome.