当前共找到 1276 篇文献分享,本页显示第 921 - 940 篇。
921.
张德祥
(2022-09-16 09:57):
#paper DOI:https://doi.org/10.1016/j.ijar.2021.09.012 Strudel: A fast and accurate learner of structured-decomposable probabilistic circuits
Probabilistic circuits (PCs)将概率分布表示为计算图,并添加图结构属性保证推理计算效率。
结构化可分解是一个吸引人的属性。
它能够有效和精确地计算复杂逻辑公式的概率,并可用于在缺失数据的情况下推理某些预测模型的预期输出。
本文提出一种简单、快速、准确的结构化可分解 PCs 学习算法 Strudel: STRUctured-DEcomposable Learner,从数据中直接学习概率计算图网络。
Abstract:
Probabilistic circuits (PCs) represent a probability distribution as a computational graph. Enforcing structural properties on these graphs guarantees that several inference scenarios become tractable. Among these properties, structured decomposability is …
>>>
Probabilistic circuits (PCs) represent a probability distribution as a computational graph. Enforcing structural properties on these graphs guarantees that several inference scenarios become tractable. Among these properties, structured decomposability is a particularly appealing one: it enables the efficient and exact computations of the probability of complex logical formulas, and can be used to reason about the expected output of certain predictive models under missing data. This paper proposes Strudel, a simple, fast and accurate learning algorithm for structured-decomposable PCs. Compared to prior work for learning structured-decomposable PCs, Strudel delivers more accurate single PC models in fewer iterations, and dramatically scales learning when building ensembles of PCs. It achieves this scalability by exploiting another structural property of PCs, called determinism, and by sharing the same computational graph across mixture components. We show these advantages on standard density estimation benchmarks and challenging inference scenarios.
<<<
翻译
922.
张德祥
(2022-09-16 09:36):
#paper
URL: http://starai.cs.ucla.edu/papers/ProbCirc20.pdf
Probabilistic circuits: A unifying framework for tractable probabilistic models
概率模型是现代机器学习(ML)和人工智能(AI)的核心。
事实上,概率论为在不确定性存在的情况下做出决策提供了一个原则性的、几乎普遍采用的机制。例如,在机器学习中,我们假设我们的数据来自未知的概率分布;
许多机器学习任务简化为简单地执行概率推理。类似地,许多形式的基于模型的人工智能寻求直接将支配我们周围世界的机制表示为某种形式的概率分布。
难怪 ML 中的许多注意力都放在从数据中学习分布上。我们将越来越多的表达性概率模型作为密度估计器,这些模型越来越接近产生数据的分布
但是之前模型及深度学习中的模型效率都不高,而且还不准确,我们要开发理论上可靠的模型且推理时间可控。且富有表现力,现在可以用统一的模型Probabilistic Circuits来处理。
Probabilistic Circuits特点:Probabilistic Circuits就是神经网络,而且是分层混合网络模型
2020.
Abstract:
No abstract available.
923.
颜林林
(2022-09-15 22:35):
#paper doi:10.1002/humu.24455 Human Mutation, 2022, de novo variant calling identifies cancer mutation signatures in the 1000 Genomes Project. 本文开发了一种能利用GPU加速、基于trio(一家三口,父母两人及一个子女)全基因组测序数据、检测新发突变(de novo variant)的工具。并使用该工具重新分析了三个大规模trio人群数据,三个人群分别是Simons Simplex Collection(SSC)、Simons Foundation Powering Autism Research(SPARK)和千人基因组(1000 Genomes Project,1000G),其样本类型分别为外周血、唾液和细胞系。结果发现细胞系的新发突变数量和特征,明显不符合预期。通过对1000G中的这些新发突变的特征分析,发现它们与B细胞淋巴瘤相似,从而推断其大多应为细胞系制备过程(即EBV处理)中引入的artifacts。
Abstract:
Detection of de novo variants (DNVs) is critical for studies of disease-related variation and mutation rates. To accelerate DNV calling, we developed a graphics processing units-based workflow. We applied our …
>>>
Detection of de novo variants (DNVs) is critical for studies of disease-related variation and mutation rates. To accelerate DNV calling, we developed a graphics processing units-based workflow. We applied our workflow to whole-genome sequencing data from three parent-child sequenced cohorts including the Simons Simplex Collection (SSC), Simons Foundation Powering Autism Research (SPARK), and the 1000 Genomes Project (1000G) that were sequenced using DNA from blood, saliva, and lymphoblastoid cell lines (LCLs), respectively. The SSC and SPARK DNV callsets were within expectations for number of DNVs, percent at CpG sites, phasing to the paternal chromosome of origin, and average allele balance. However, the 1000G DNV callset was not within expectations and contained excessive DNVs that are likely cell line artifacts. Mutation signature analysis revealed 30% of 1000G DNV signatures matched B-cell lymphoma. Furthermore, we found variants in DNA repair genes and at Clinvar pathogenic or likely-pathogenic sites and significant excess of protein-coding DNVs in IGLL5; a gene known to be involved in B-cell lymphomas. Our study provides a new rapid DNV caller for the field and elucidates important implications of using sequencing data from LCLs for reference building and disease-related projects.
<<<
翻译
924.
颜林林
(2022-09-14 05:52):
#paper doi:10.1002/humu.24460 Human Mutation, 2022, CIC missense variants contribute to susceptibility for spina bifida. 既往研究发现,叶酸摄入对于神经系统发育具有重要作用,其缺乏可能导致神经管缺陷(Neural tube defects,NTDs)这样的严重先天畸形。本文应该是从另一项研究出发,由入组的140例散发脊柱裂(spina bifida)病例,进行的全基因组测序结果中,发现8例CIC基因的罕见错义突变。通过近缘物种间序列保守性,确认了这些突变可能存在重要作用。在细胞系中通过质粒转染和叶酸缺乏培养等实验,引入野生型或携带上述突变的CIC基因的质粒,通过免疫荧光观察突变对表达量和亚细胞定位的影响。此外,还使用Western、qPCR等方法,对CIC所调控的基因的表达进行测定,确认了所发现的CIC突变,确实会对相关通路造成影响。这是一篇用湿实验方法对所发现基因突变功能进行验证的典型研究。
Abstract:
Neural tube defects (NTDs) are congenital malformations resulting from abnormal embryonic development of the brain, spine, or spinal column. The genetic etiology of human NTDs remains poorly understood despite intensive …
>>>
Neural tube defects (NTDs) are congenital malformations resulting from abnormal embryonic development of the brain, spine, or spinal column. The genetic etiology of human NTDs remains poorly understood despite intensive investigation. CIC, homolog of the Capicua transcription repressor, has been reported to interact with ataxin-1 (ATXN1) and participate in the pathogenesis of spinocerebellar ataxia type 1. Our previous study demonstrated that CIC loss of function (LoF) variants contributed to the cerebral folate deficiency syndrome by downregulating folate receptor 1 (FOLR1) expression. Given the importance of folate transport in neural tube formation, we hypothesized that CIC variants could contribute to increased risk for NTDs by depressing embryonic folate concentrations. In this study, we examined CIC variants from whole-genome sequencing (WGS) data of 140 isolated spina bifida cases and identified eight missense variants of CIC gene. We tested the pathogenicity of the observed variants through multiple in vitro experiments. We determined that CIC variants decreased the FOLR1 protein level and planar cell polarity (PCP) pathway signaling in a human cell line (HeLa). In a murine cell line (NIH3T3), CIC loss of function variants downregulated PCP signaling. Taken together, this study provides evidence supporting CIC as a risk gene for human NTD.
<<<
翻译
925.
颜林林
(2022-09-13 07:21):
#paper doi:10.1016/j.vaccine.2022.08.036 Vaccine, 2022, Serious adverse events of special interest following mRNA COVID-19 vaccination in randomized trials in adults. 这篇文章跟进了Pfizer和Moderna两家公司的新冠RNA疫苗的三期临床试验,针对其报出的严重不良反应进行二次分析,确认各自疫苗相对于安慰剂所增加的风险比值。该结果提示应该进行更加详尽正式的利弊分析。而文末也再次呼吁要求公开受试者级别的相关数据,以保证临床试验的透明度和各类评估分析得以正确进行。
Abstract:
INTRODUCTION: In 2020, prior to COVID-19 vaccine rollout, the Brighton Collaboration created a priority list, endorsed by the World Health Organization, of potential adverse events relevant to COVID-19 vaccines. We …
>>>
INTRODUCTION: In 2020, prior to COVID-19 vaccine rollout, the Brighton Collaboration created a priority list, endorsed by the World Health Organization, of potential adverse events relevant to COVID-19 vaccines. We adapted the Brighton Collaboration list to evaluate serious adverse events of special interest observed in mRNA COVID-19 vaccine trials.METHODS: Secondary analysis of serious adverse events reported in the placebo-controlled, phase III randomized clinical trials of Pfizer and Moderna mRNA COVID-19 vaccines in adults (NCT04368728 and NCT04470427), focusing analysis on Brighton Collaboration adverse events of special interest.RESULTS: Pfizer and Moderna mRNA COVID-19 vaccines were associated with an excess risk of serious adverse events of special interest of 10.1 and 15.1 per 10,000 vaccinated over placebo baselines of 17.6 and 42.2 (95 % CI -0.4 to 20.6 and -3.6 to 33.8), respectively. Combined, the mRNA vaccines were associated with an excess risk of serious adverse events of special interest of 12.5 per 10,000 vaccinated (95 % CI 2.1 to 22.9); risk ratio 1.43 (95 % CI 1.07 to 1.92). The Pfizer trial exhibited a 36 % higher risk of serious adverse events in the vaccine group; risk difference 18.0 per 10,000 vaccinated (95 % CI 1.2 to 34.9); risk ratio 1.36 (95 % CI 1.02 to 1.83). The Moderna trial exhibited a 6 % higher risk of serious adverse events in the vaccine group: risk difference 7.1 per 10,000 (95 % CI -23.2 to 37.4); risk ratio 1.06 (95 % CI 0.84 to 1.33). Combined, there was a 16 % higher risk of serious adverse events in mRNA vaccine recipients: risk difference 13.2 (95 % CI -3.2 to 29.6); risk ratio 1.16 (95 % CI 0.97 to 1.39).DISCUSSION: The excess risk of serious adverse events found in our study points to the need for formal harm-benefit analyses, particularly those that are stratified according to risk of serious COVID-19 outcomes. These analyses will require public release of participant level datasets.
<<<
翻译
926.
颜林林
(2022-09-11 23:59):
#paper doi:10.1101/2022.09.09.453067 bioRxiv, 2022, HexSE: Simulating evolution in overlapping reading frames. 重叠基因是在病毒(质粒)中发现的一种有趣现象,即同一段核酸序列,因为翻译蛋白质的起始位置不同(即阅读框不同)导致形成不同蛋白。到目前为止的研究,发现在许多物种中都存在此现象。本文通过分析序列演化速率,来从积累的大量已被测序的基因组数据中,寻找这样的重叠基因。其基本假设是,如果存在重叠基因,则相应序列上受到的演化选择压力会有所不同,于是在结果上呈现出不同的演化速率。这是个很有意思的思路和研究课题。
bioRxiv,
2022.
DOI: 10.1101/2022.09.09.453067
Abstract:
Motivation: Gene overlap occurs when two or more genes are encoded by the same nucleotides. This phenomenon is found in all taxonomic domains, but is particularly common in viruses, where …
>>>
Motivation: Gene overlap occurs when two or more genes are encoded by the same nucleotides. This phenomenon is found in all taxonomic domains, but is particularly common in viruses, where it may provide a mechanism to increase the information content of compact genomes. The presence of overlapping reading frames (OvRFs) can skew estimates of selection based on the rates of non-synonymous and synonymous substitutions, since a substitution that is synonymous in one reading frame may be non-synonymous in another, and vice versa. Results: To understand the impact of OvRFs on molecular evolution, we implemented a versatile simulation model of nucleotide sequence evolution along a phylogeny with an arbitrary distribution of reading frames. We use a custom data structure to track the substitution rates at every nucleotide site, which is determined by the stationary nucleotide frequencies, transition bias, and the distribution of selection biases (dN/dS) in the respective reading frames. Availability and implementation: Our simulation model is implemented in the Python scripting language. All source code is released under the GNU General Public License (GPL) version 3, and is available at https://github.com/PoonLab/HexSE.
<<<
翻译
927.
song
(2022-09-09 09:04):
#paper https://doi.org/10.48550/arXiv.2206.13236 Pruned RNN-T for fast, memory-efficient ASR training
来自于小米新一代kaldi团队。RNN-T是目前端到端语音识别的主流范式之一,是目前流式解码模型中表现最好和最易工业化部署的,缺点是训练时内存比其他主流模型占用内存至少高一个数量级。究其原因是因为比其他模型如CTC和attention模型的内存多了一个解码器的输出帧数,U,导致的。U值一般在几十到几百之间。本文提出了一种在不降低模型性能的情况下对模型进行剪枝以降低U值的方法。该团队首先发现在RNN-T loss计算过程中,并不是每个计算节点都参与进了计算过程中。计算节点的数量和输出帧数U成正比,只要选择并只保留对模型训练有作用的计算节点便可减少模型内存提高模型训练速度。在计算梯度过程中,只有中间一段连续的计算节点参与进训练之中,根据不同的常见,这个连续节点数,S,为4或5。在实验中,训练时间达到之前sota的约十六分之一,内存占用达到之前的约五分之一,模型性能仅降了0.05%。个人尝试下来,仅用4张V100已经较少的调参便可完全重现并部署。中小型公司将sota模型应用于产品之中的成本和人力将大大减少
arXiv,
2022.
DOI: 10.48550/arXiv.2206.13236
Abstract:
The RNN-Transducer (RNN-T) framework for speech recognition has been growing in popularity, particularly for deployed real-time ASR systems, because it combines high accuracy with naturally streaming recognition. One of the …
>>>
The RNN-Transducer (RNN-T) framework for speech recognition has been growing in popularity, particularly for deployed real-time ASR systems, because it combines high accuracy with naturally streaming recognition. One of the drawbacks of RNN-T is that its loss function is relatively slow to compute, and can use a lot of memory. Excessive GPU memory usage can make it impractical to use RNN-T loss in cases where the vocabulary size is large: for example, for Chinese character-based ASR. We introduce a method for faster and more memory-efficient RNN-T loss computation. We first obtain pruning bounds for the RNN-T recursion using a simple joiner network that is linear in the encoder and decoder embeddings; we can evaluate this without using much memory. We then use those pruning bounds to evaluate the full, non-linear joiner network.
<<<
翻译
928.
魏魏魏
(2022-09-07 11:47):
#paper doi:10.1007/s10802-010-9396-z. Journal of Abnormal Child Psychology, (2010), Mother and Adolescent Reports of Associations Between Child Behavior Problems and Mother-Child Relationship Qualities: Separating Shared Variance from Individual Variance. 基于共同命运模型(Common Fate Model, CFM)的研究很少,所以看到了2010年的文献,只为更好地学习这种方法。共同命运模型很适合研究夫妻、母子和父子关系这种双方成员共享生活环境的人,即双方受到共同的环境变量影响,在一些变量上双方具有相似性。关系中的双方都需要在相关变量上报告自己的情况,这样就形成了配对数据(dyadic data),而且,双方的数据会存在依存性(interdependence),这也就打破了传统的相关分析需要变量各自独立的假设前提,此时共同命运模型可以解决这个问题。再有,传统研究只考察了单个被试在自变量和结果变量上的情况,这可能会出现因数据有共同来源而导致的共同方法变异(Common method variance),这会使最终结果的变异被夸大或缩小,也会影响我们对实际情况的准确认识。此时,这个模型也很有优势,因为它引进了另一个关系被试的情况,使得数据的来源多元化。基于共同命运模型的分析除了考察单个被试内变量的相关情况,也考察了被试间在同样的变量上的相关情况,还考察了关系水平上自变量与结果变量的相关情况。在这个过程中,关系双方共享因素带来的变异被分解了出来,帮助人们更好地了解了自变量与因变量的真实关系。当前研究考察了青少年行为问题与母子关系品质的关系,在两个变量上,母子双方有共同的认识,彼此间也会存在差异。基于共同命运模型,该研究同时考察了母子在相同变量上的情况。在具体分析中,除了考察子女报告的变量间的相关情况,也考察了母亲报告的变量间的相关情况,还从母子关系水平上分析了变量间的相关,并同时在模型中分别分析了母子在变量间的相关情况,并比较了多个相关系数之间在大小上的差异情况。最终发现了不同于基于传统研究的发现结果。
Abstract:
This study contrasts results from different correlational methods for examining links between mother and child (N = 72 dyads) reports of early adolescent (M = 11.5 years) behavior problems and …
>>>
This study contrasts results from different correlational methods for examining links between mother and child (N = 72 dyads) reports of early adolescent (M = 11.5 years) behavior problems and relationship negativity and support. Simple (Pearson) correlations revealed a consistent pattern of statistically significant associations, regardless of whether scores came from the same reporter or from different reporters. When correlations between behavior problems and relationship quality differed, within-reporter correlations were always greater in magnitude than between-reporter correlations. Dyadic (common fate) analyses designed for interdependent data decomposed within-reporter correlations into variance shared across reporters (dyadic correlations) and variance unique to specific reporters (individual correlations). Dyadic correlations were responsible for most associations between adolescent behavior problems and relationship negativity; after partitioning variance shared across reporters, no individual correlations emerged as statistically significant. In contrast, adolescent behavior problems were linked to relationship support via both shared variance and variance unique to maternal perceptions. Dyadic analyses provide a parsimonious alternative to multiple contrasts in instances when identical measures have been collected from multiple reporters. Findings from these analyses indicate that same-reporter variance bias should not be assumed in the absence of dyadic statistical analyses.
<<<
翻译
929.
马斯克齊
(2022-09-04 21:57):
#paper doi:10.19695/j.cnki.cn12-1369,2022,论人工智能在大学校园的重要应用。随着人工智能技术的不断发展以及疫情大背景下,校园学习如何与时代接轨列出人工智能在校园的一些应用,智能教学,智慧图书馆,智慧校园生活等都会有不一样的体验,同时对未来的教学改革产生重要影响。
数字技术与应用,
2022.
DOI: 10.19695/j.cnki.cn12-1369.2022.07.22
Abstract:
近年来,随着全球经济的快速发展,计算机科学技术的迅猛进步和发展以及高校智慧管理对于科学技术应用的迫切需求,物联网应用、大数据管理、5G和云计算等前言技术都逐渐在高校智慧管理上得到了体现和普及。而人工智能技术作为当前在高校管理中最受青睐的高科技技术,已在教学研究管理、在线教育和学生校园生活优化等各个方面得到了深度应用并在取得显著成效。本文介绍人工智能现阶段在高校校园管理中的重要应用,以供参考。
>>>
近年来,随着全球经济的快速发展,计算机科学技术的迅猛进步和发展以及高校智慧管理对于科学技术应用的迫切需求,物联网应用、大数据管理、5G和云计算等前言技术都逐渐在高校智慧管理上得到了体现和普及。而人工智能技术作为当前在高校管理中最受青睐的高科技技术,已在教学研究管理、在线教育和学生校园生活优化等各个方面得到了深度应用并在取得显著成效。本文介绍人工智能现阶段在高校校园管理中的重要应用,以供参考。
<<<
翻译
930.
张德祥
(2022-09-01 22:03):
#paper https://doi.org/10.48550/arXiv.2208.11970 Understanding Diffusion Models: A Unified Perspective ;最近大火的视频生成模型 dall-e 等背后都是diffusion 模型,这篇论文细致的讲解了diffusion模型的来龙去脉,从ELBO 到VAE 到hierarchical VAE 到diffusion 模型,及diffusion模型的三个视角及diffusion模型的局限,整篇论文公式推导清晰易读是了解diffusion模型的好资料。
arXiv,
2022.
DOI: 10.48550/arXiv.2208.11970
Abstract:
Diffusion models have shown incredible capabilities as generative models; indeed, they power the current state-of-the-art models on text-conditioned image generation such as Imagen and DALL-E 2. In this work we …
>>>
Diffusion models have shown incredible capabilities as generative models; indeed, they power the current state-of-the-art models on text-conditioned image generation such as Imagen and DALL-E 2. In this work we review, demystify, and unify the understanding of diffusion models across both variational and score-based perspectives. We first derive Variational Diffusion Models (VDM) as a special case of a Markovian Hierarchical Variational Autoencoder, where three key assumptions enable tractable computation and scalable optimization of the ELBO. We then prove that optimizing a VDM boils down to learning a neural network to predict one of three potential objectives: the original source input from any arbitrary noisification of it, the original source noise from any arbitrarily noisified input, or the score function of a noisified input at any arbitrary noise level. We then dive deeper into what it means to learn the score function, and connect the variational perspective of a diffusion model explicitly with the Score-based Generative Modeling perspective through Tweedie's Formula. Lastly, we cover how to learn a conditional distribution using diffusion models via guidance.
<<<
翻译
931.
魏魏魏
(2022-09-01 15:01):
#paper doi:10.1007/978-3-319-21635-5_9 National Symposium on Family Issues, (2016), Why Do Fathers Matter for Children’s Development?
延续此前阅读的内容,持续关注父亲和父亲的教养行为对子女发展的影响。这是一篇评论或综述性文章,作者介绍了此前研究中发现的父亲不同于母亲的教养行为,比如参与的行为领域不同,他更喜欢打闹类的游戏,而母亲更多是照料行为等,而且,发现这种不同对子女发展的影响也是有差异的。作者对此做出了分析和点评,认为这可能都是表面现象,可能都受到了文化影响下的教养观念的影响,比如,子女性别会影响父亲具体的教养行为。他认为,父母教养行为可能在本质上可能区别并不大。作者认为,为了更好地理解父亲的教养行为对子女的影响,可以基于“父子关系生态学的扩展模型”来分析相关影响因素(环境因素和个人特征因素)以及具体的影响方式。以后的研究应该继续关注父亲的教养行为并深入分析其具体机制。
.
Abstract:
Evidence regarding whether fathers are “essential” for or contribute substantially to children’s development over and above their financial contributions is still emerging. The question of how exactly fathers matter for …
>>>
Evidence regarding whether fathers are “essential” for or contribute substantially to children’s development over and above their financial contributions is still emerging. The question of how exactly fathers matter for children’s development has dominated much of the emerging research on fatherhood and has produced some promising findings. In this chapter, I briefly review the theoretical underpinnings of what fathers do and how it matters for children’s development. I then examine current findings on the association between fathers’ involvement and children’s outcomes over time, and conclude with suggestions for future research.
<<<
翻译
932.
王昊
(2022-09-01 14:36):
#paper doi:10.1109/TNNLS.2022.3152527 Hwanjun Song, Minseok Kim, Dongmin Park, Yooju Shin, and Jae-Gil Lee. 2022. Learning From Noisy Labels With Deep Neural Networks: A Survey. IEEE Transactions on Neural Networks and Learning Systems: 1–19. 本文是噪声标签2022年的综述论文,主要介绍结构化数据、图片分类数据集等进行去噪的各种方法。具体类别总结如下:
【Robust Architecture】基于attention注意力机制给干净样本和噪声数据进行打分,文章叫做Attention Feature Mixup,在计算最终loss的时候有两部分,一部分是同一个类的每张图和标签计算的交叉熵损失;另外一个损失是数据mixup得到的新的数据x'和标签y'计算的loss.
【Robust Regularization】
这一部分主要是通过一些添加正则ticks,防止模型过拟合到噪声数据上,常用的正则方法包含:label smooth、l1、l2、MixUp等.
【Sample Selection】Area Under the Margin metric (AUM):在训练过程中一边训练一边筛选数据的方式.
【数据划分】是通过密度聚类的思路,将一个类的数据分成easy dataset、smi-hard dataset 和 hard dataset,一般噪声数据是较为困难训练的数据,对于每张图分配一个权重,文中建议1.0、0.5和0.5;模型的训练借鉴了课程学习的思路.
【Semi-supervised Learning】基于半监督学习的带噪学习算法,首先介绍DivideMix方法,其实还是co-teaching的思路,但是在挑出干净样本和噪音样本后,把噪音样本当做无标签样本,通过 FixMatch 的方法进行训练,目前半监督图像分类的 SOTA 应该还是 FixMatch. (这个性能比较好)
【Label correction】“label correction phase”通过一个pre-trained模型得到随机选择每个类中的几张图采用聚类的方法得到Prototype样本的每个类的聚类中心,对输入图片得到的特征向量和各类聚类中心计算距离,得到图片的伪标签,最后的loss是原始标签计算的交叉熵损失和伪标签计算的伪标签的求和。
IF:10.200Q1
IEEE transactions on neural networks and learning systems,
2023-Nov.
DOI: 10.1109/TNNLS.2022.3152527
PMID: 35254993
Abstract:
Deep learning has achieved remarkable success in numerous domains with help from large amounts of big data. However, the quality of data labels is a concern because of the lack …
>>>
Deep learning has achieved remarkable success in numerous domains with help from large amounts of big data. However, the quality of data labels is a concern because of the lack of high-quality labels in many real-world scenarios. As noisy labels severely degrade the generalization performance of deep neural networks, learning from noisy labels (robust training) is becoming an important task in modern deep learning applications. In this survey, we first describe the problem of learning with label noise from a supervised learning perspective. Next, we provide a comprehensive review of 62 state-of-the-art robust training methods, all of which are categorized into five groups according to their methodological difference, followed by a systematic comparison of six properties used to evaluate their superiority. Subsequently, we perform an in-depth analysis of noise rate estimation and summarize the typically used evaluation methodology, including public noisy datasets and evaluation metrics. Finally, we present several promising research directions that can serve as a guideline for future studies.
<<<
翻译
933.
王昊
(2022-09-01 14:34):
#paper doi:10.1109/ICCV48922.2021.00014 ZHOU X, LIU X, WANG C, 等. Learning with Noisy Labels via Sparse Regularization[C/OL]//2021 IEEE/CVF International Conference on Computer Vision (ICCV). 2021: 72-81. https://doi.org/10.1109/ICCV48922.2021.00014. 本文使用稀疏正则化的方法,将输出尽可能地往one-hot上引导,使得输出锐化(一个是1,其它都是0,相当于有很大的确信度就是那一个答案,其它的概率都很低), 具体使用使用Lp Norm方法来达成. 该方法属于噪声标签去噪的损失函数方法的paper。噪声标签去噪综述可参见: SONG H, KIM M, PARK D, 等. Learning From Noisy Labels With Deep Neural Networks: A Survey[J/OL]. IEEE Transactions on Neural Networks and Learning Systems, 2022: 1-19. https://doi.org/10.1109/TNNLS.2022.3152527
Abstract:
Learning with noisy labels is an important and challenging task for training accurate deep neural networks. Some commonly-used loss functions, such as Cross Entropy (CE), suffer from severe overfitting to …
>>>
Learning with noisy labels is an important and challenging task for training accurate deep neural networks. Some commonly-used loss functions, such as Cross Entropy (CE), suffer from severe overfitting to noisy labels. Robust loss functions that satisfy the symmetric condition were tailored to remedy this problem, which however encounter the underfitting effect. In this paper, we theoretically prove that any loss can be made robust to noisy labels by restricting the network output to the set of permutations over a fixed vector. When the fixed vector is one-hot, we only need to constrain the output to be one-hot, which however produces zero gradients almost everywhere and thus makes gradient-based optimization difficult. In this work, we introduce the sparse regularization strategy to approximate the one-hot constraint, which is composed of network output sharpening operation that enforces the output distribution of a net-work to be sharp and the ℓ p -norm (p ≤ 1) regularization that promotes the network output to be sparse. This simple approach guarantees the robustness of arbitrary loss functions while not hindering the fitting ability. Experimental results demonstrate that our method can significantly improve the performance of commonly-used loss functions in the presence of noisy labels and class imbalance, and out-perform the state-of-the-art methods. The code is available at https://github.com/hitcszx/lnl_sr.
<<<
翻译
934.
龙海晨
(2022-09-01 12:18):
#paper Luo F, Lin K. N6-methyladenosine (m6A) reader IGF2BP1 accelerates gastric cancer aerobic glycolysis in c-Myc-dependent manner[J]. Experimental Cell Research, 2022, 417(1): 113176. PMID: 35489385
DOI: 10.1016/j.yexcr.2022.113176
该文章研究了N 6-甲基腺苷(m 6 A),RNA m 6 A reader IGF2BP1 在胃癌发生发展中的作用。m 6 A参与调节多种癌症中的癌细胞增殖、转移、形成。文章分析了IGF2BP1在胃癌(gastric cancer,GC)中的功能和机制。研究结果表明,IGF2BP1 在 GC 组织中上调,可作为 GC 患者预后不良的预测因子。在功能上,IGF2BP1 促进体外 GC 细胞迁移和有氧糖酵解。IGF2BP1 敲低抑制了体内肿瘤的生长。文章证明了 IGF2BP1 通过 m6A 依赖性方式直接与 c-MYC mRNA 相互作用。文章发现,m 6 A reader IGF2BP1 以依赖 m 6 A/c-Myc 方式促进了 GC 的致癌性,这可能为 GC 提供治疗策略。
Abstract:
The N-methyladenosine (mA) is involved in the regulation of cell proliferation and metastasis formation in multiple cancers. However, the biological significance of RNA mA reader IGF2BP1 and the modification of …
>>>
The N-methyladenosine (mA) is involved in the regulation of cell proliferation and metastasis formation in multiple cancers. However, the biological significance of RNA mA reader IGF2BP1 and the modification of IGF2BP1 itself have not been fully investigated. Here, we analyzed the functions and mechanism of IGF2BP1 in gastric cancer (GC). Results showed that IGF2BP1 upregulated in GC tissue and acted as a predictor of poor prognosis for GC patients. Functionally, IGF2BP1 promoted the migration and aerobic glycolysis of GC cells in vitro. Moreover, IGF2BP1 knockdown repressed the tumor growth in vivo. We also demonstrated that IGF2BP1 directly interacted with c-MYC mRNA via m6A-dependent manner to by stabilize its stability. Overall, these findings demonstrated that mA reader IGF2BP1 facilitated the carcinogenic of GC in mA/c-Myc-dependent manner, which might provide critical therapeutic strategy for GC.
<<<
翻译
935.
小擎子
(2022-08-31 23:41):
#paper doi: 10.1126/science.aaw7479 Nature, 2022, Akkermansia muciniphila phospholipid induces homeostatic immune responses. 找到了明星益生菌Akk调节免疫的物质a15:0-i15:0 PE,解释了其作用机制。a15:0-i15:0 PE是Akk菌膜脂质的主要成分之一。Akk菌隶属于疣微菌门(Verrucomicrobia),是肠道微生物中唯一一个来自疣微菌门的细菌。大海捞针式的研究,对什么样的生长环境能使Akk菌产生a15:0-i15:0 PE比较好奇。
Abstract:
Intestinal adaptive immune responses influence host health, yet only a few intestinal bacteria species that induce cognate adaptive immune responses during homeostasis have been identified. Here, we show that , …
>>>
Intestinal adaptive immune responses influence host health, yet only a few intestinal bacteria species that induce cognate adaptive immune responses during homeostasis have been identified. Here, we show that , an intestinal bacterium associated with systemic effects on host metabolism and PD-1 checkpoint immunotherapy, induces immunoglobulin G1 (IgG1) antibodies and antigen-specific T cell responses in mice. Unlike previously characterized mucosal responses, T cell responses to are limited to T follicular helper cells in a gnotobiotic setting, without appreciable induction of other T helper fates or migration to the lamina propria. However, -specific responses are context dependent and adopt other fates in conventional mice. These findings suggest that, during homeostasis, contextual signals influence T cell responses to the microbiota and modulate host immune function.
<<<
翻译
936.
洪媛媛
(2022-08-31 23:32):
#paper http//10.2144/97233rr01 BioTechniques 23:504-511 (1997). Multiplex PCR: Critical Parameters and Step-by-Step Protocol. 介绍了多重PCR的优化方法,针对不同的异常情形给出了对应的优化措施,并且讲解了如何从延伸温度、延伸时间、退火时间和温度、PCR循环数、引物浓度、dNTP和Mg离子浓度、PCR buffer中KCl浓度、DNA和酶量和添加剂比如DMSO等方面进行优化的技术细节。
Abstract:
By simultaneously amplifying more than one locus in the same reaction, multiplex PCR is becoming a rapid and convenient screening assay in both the clinical and the research laboratory. While …
>>>
By simultaneously amplifying more than one locus in the same reaction, multiplex PCR is becoming a rapid and convenient screening assay in both the clinical and the research laboratory. While numerous papers and manuals discuss in detail conditions influencing the quality of PCR in general, relatively little has been published about the important experimental factors and the common difficulties frequently encountered with multiplex PCR. We have examined various conditions of the multiplex PCR, using a large number of primer pairs. Especially important for a successful multiplex PCR assay are the relative concentrations of the primers at the various loci, the concentration of the PCR buffer, the cycling temperatures and the balance between the magnesium chloride and deoxynucleotide concentrations. Based on our experience, we propose a protocol for developing a multiplex PCR assay and suggest ways to overcome commonly encountered problems.
<<<
翻译
937.
prayer
(2022-08-31 23:27):
#paper DOI:10.7326/M22-0924
The Association of Baseline Plasma SARS-CoV-2 Nucleocapsid Antigen Level and Outcomes in Patients Hospitalized With COVID-19
这是一篇多中心临床研究,纳入了2540位参与者,探究发现SARS-CoV-2基线抗原水平与肺部症状严重性、住院时长等密切相关。可影响血浆抗原水平的因素包括抗spike抗体水平、男性、肾损伤等。血浆抗原水平可作为COVID-19住院患者实用且具有临床意义的biomarker。
Abstract:
BACKGROUND: Levels of plasma SARS-CoV-2 nucleocapsid (N) antigen may be an important biomarker in patients with COVID-19 and enhance our understanding of the pathogenesis of COVID-19.OBJECTIVE: To evaluate whether levels …
>>>
BACKGROUND: Levels of plasma SARS-CoV-2 nucleocapsid (N) antigen may be an important biomarker in patients with COVID-19 and enhance our understanding of the pathogenesis of COVID-19.OBJECTIVE: To evaluate whether levels of plasma antigen can predict short-term clinical outcomes and identify clinical and viral factors associated with plasma antigen levels in hospitalized patients with SARS-CoV-2.DESIGN: Cross-sectional study of baseline plasma antigen level from 2540 participants enrolled in the TICO (Therapeutics for Inpatients With COVID-19) platform trial from August 2020 to November 2021, with additional data on day 5 outcome and time to discharge.SETTING: 114 centers in 10 countries.PARTICIPANTS: Adults hospitalized for acute SARS-CoV-2 infection with 12 days or less of symptoms.MEASUREMENTS: Baseline plasma viral N antigen level was measured at a central laboratory. Delta variant status was determined from baseline nasal swabs using reverse transcriptase polymerase chain reaction. Associations between baseline patient characteristics and viral factors and baseline plasma antigen levels were assessed using both unadjusted and multivariable modeling. Association between elevated baseline antigen level of 1000 ng/L or greater and outcomes, including worsening of ordinal pulmonary scale at day 5 and time to hospital discharge, were evaluated using logistic regression and Fine-Gray regression models, respectively.RESULTS: Plasma antigen was below the level of quantification in 5% of participants at enrollment, and 1000 ng/L or greater in 57%. Baseline pulmonary severity of illness was strongly associated with plasma antigen level, with mean plasma antigen level 3.10-fold higher among those requiring noninvasive ventilation or high-flow nasal cannula compared with room air (95% CI, 2.22 to 4.34). Plasma antigen level was higher in those who lacked antispike antibodies (6.42 fold; CI, 5.37 to 7.66) and in those with the Delta variant (1.73 fold; CI, 1.41 to 2.13). Additional factors associated with higher baseline antigen level included male sex, shorter time since hospital admission, decreased days of remdesivir, and renal impairment. In contrast, race, ethnicity, body mass index, and immunocompromising conditions were not associated with plasma antigen levels. Plasma antigen level of 1000 ng/L or greater was associated with a markedly higher odds of worsened pulmonary status at day 5 (odds ratio, 5.06 [CI, 3.41 to 7.50]) and longer time to hospital discharge (median, 7 vs. 4 days; subhazard ratio, 0.51 [CI, 0.45 to 0.57]), with subhazard ratios similar across all levels of baseline pulmonary severity.LIMITATIONS: Plasma samples were drawn at enrollment, not hospital presentation. No point-of-care test to measure plasma antigen is currently available.CONCLUSION: Elevated plasma antigen is highly associated with both severity of pulmonary illness and clinically important patient outcomes. Multiple clinical and viral factors are associated with plasma antigen level at presentation. These data support a potential role of ongoing viral replication in the pathogenesis of SARS-CoV-2 in hospitalized patients.PRIMARY FUNDING SOURCE: U.S. government Operation Warp Speed and National Institute of Allergy and Infectious Diseases.
<<<
翻译
938.
Spring
(2022-08-31 23:17):
#paper doi: 10.1038/s41590-022-01284-1
Inflammation triggers ILC3 patrolling of the intestinal barrier
① 通过活体成像,观察到小鼠肠道绒毛处的ILC3大多处于几乎不移动的稳定状态;② 利用细菌鞭毛蛋白诱导炎症发生,可增加ILC3的IL-22表达,并增强ILC3的移动能力,ILC3在肠道中的迁移模式也发生显著改变;③ T细胞可抑制ILC3的肠道巡逻行为,而CCR9-CCL25趋化因子信号可增强炎症诱导的ILC3迁移;④ 阻断CCL25介导的ILC3迁移可促进肠道上皮细胞的死亡,从而破坏肠道屏障,ILC3产生的IL-22可抑制肠道上皮细胞的死亡。
Abstract:
An orchestrated cellular network, including adaptive lymphocytes and group 3 innate lymphoid cells (ILC3s), maintains intestinal barrier integrity and homeostasis. T cells can monitor environmental insults through constitutive circulation, scanning …
>>>
An orchestrated cellular network, including adaptive lymphocytes and group 3 innate lymphoid cells (ILC3s), maintains intestinal barrier integrity and homeostasis. T cells can monitor environmental insults through constitutive circulation, scanning tissues and forming immunological contacts, a process named immunosurveillance. In contrast, the dynamics of intestinal ILC3s are unknown. Using intravital imaging, we observed that villus ILC3s were largely immotile at steady state but acquired migratory 'patrolling' attributes and enhanced cytokine expression in response to inflammation. We showed that T cells, the chemokine CCL25 and bacterial ligands regulated intestinal ILC3 behavior and that loss of patrolling behavior by interleukin-22 (IL-22)-producing ILC3s altered the intestinal barrier through increased epithelial cell death. Collectively, we identified notable differences between the behavior of ILC3s and T cells, with a prominent adaptation of intestinal ILC3s toward mucosal immunosurveillance after inflammation.
<<<
翻译
939.
符毓
(2022-08-31 23:08):
#paper doi: 10.1525/cmr.2014.57.1.16 California Management Review, 2014, The Japanese software industry: what went wrong and what can we learn from it?
工业软件总是伴随工业制造一起发展,然而日本作为全球领先的先进制造大国,日产工业软件明显缺失。本文提出了几个解释:
1. 软件高端人才缺失:重硬件轻软件的文化加上软件过度外包和定制化,让软件人才“蓝领化”
2. 创业公司缺失:日本常见的企业形态是由老牌大型公司开拓新产品新业务,再分拆出子公司,新公司沿袭老公司的员工和制度,并不会挑战老公司;与之类似,当新供应商提供新产品或者服务的时候,客户通常会等待甚至支持原有供应商提供类似服务,而不是切换供应商
Abstract:
Recent findings indicate that the Japanese IT sector increasingly lags the U.S. IT sector in software innovation and that this underlies Japan's weakening competitive performance vis-à-vis U.S. IT. This article …
>>>
Recent findings indicate that the Japanese IT sector increasingly lags the U.S. IT sector in software innovation and that this underlies Japan's weakening competitive performance vis-à-vis U.S. IT. This article explores alternative explanations for this outcome and analyzes what explains the Japanese software industry's trajectory. The sources are found in the late understanding of the transformational role of software and its value-creating potential as well as in the evolution of the industry's structure. Finally, this article considers what policy makers in other nations might learn from the Japanese experience in building a more vibrant software industry.
<<<
翻译
940.
林李泽强
(2022-08-31 22:57):
#paper doi:10.1093/bioinformatics/btz128,Bioinformatics,2019,Brain annotation toolbox: exploring thefunctional and genetic associations of neuroimaging results. 过去以来,大多数神经影像学研究的结果(比如激活的簇/区域或大脑区域之间的功能连接),往往无法方便和系统地解释,导致生物学意义不明确。在这项研究中,作者开发了一个大脑注释工具箱,它可以为神经成像结果自动生成功能和基因注释。该工具包是基于Neurosynth数据库中的体素级功能描述以及Allen人脑图谱中的基因表达谱,将它们用于生成区域级神经成像结果的功能/基因信息。这个工具包是基于MATLAB的免费的开源工具包,可以帮助为新发现的具有未知功能的区域提供功能/基因注释。
Abstract:
MOTIVATION: Advances in neuroimaging and sequencing techniques provide an unprecedented opportunity to map the function of brain regions and identify the roots of psychiatric diseases. However, the results from most …
>>>
MOTIVATION: Advances in neuroimaging and sequencing techniques provide an unprecedented opportunity to map the function of brain regions and identify the roots of psychiatric diseases. However, the results from most neuroimaging studies, i.e. activated clusters/regions or functional connectivities between brain regions, frequently cannot be conveniently and systematically interpreted, rendering the biological meaning unclear.RESULTS: We describe a brain annotation toolbox that generates functional and genetic annotations for neuroimaging results. The voxel-level functional description from the Neurosynth database and gene expression profile from the Allen Human Brain Atlas are used to generate functional/genetic information for region-level neuroimaging results. The validity of the approach is demonstrated by showing that the functional and genetic annotations for specific brain regions are consistent with each other; and further the region by region functional similarity network and genetic similarity network are highly correlated for major brain atlases. One application of brain annotation toolbox is to help provide functional/genetic annotations for newly discovered regions with unknown functions, e.g. the 97 new regions identified in the Human Connectome Project. Importantly, this toolbox can help understand differences between psychiatric patients and controls, and this is demonstrated using schizophrenia and autism data, for which the functional and genetic annotations for the neuroimaging changes in patients are consistent with each other and help interpret the results.AVAILABILITY AND IMPLEMENTATION: BAT is implemented as a free and open-source MATLAB toolbox and is publicly available at http://123.56.224.61:1313/post/bat.SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
<<<
翻译