分类： 计算机科学 >> 计算机应用技术 提交时间： 2017-05-20
摘要： Decoding human brain activities via functional magnetic resonance imaging (fMRI) has gained increasing attention in recent years. While encouraging results have been reported in brain states classification tasks, reconstructing the details of human visual experience still remains difficult. Two main challenges that hinder the development of effective models are the perplexing fMRI measurement noise and the high dimensionality of limited data instances. Existing methods generally suffer from one or both of these issues and yield dissatisfactory results. In this paper, we tackle this problem by casting the reconstruction of visual stimulus as the Bayesian inference of missing view in a multiview latent variable model. Sharing a common latent representation, our joint generative model of external stimulus and brain response is not only ``deep" in extracting nonlinear features from visual images, but also powerful in capturing correlations among voxel activities of fMRI recordings. The nonlinearity and deep structure endow our model with strong representation ability, while the correlations of voxel activities are critical for suppressing noise and improving prediction. We devise an efficient variational Bayesian method to infer the latent variables and the model parameters. To further improve the reconstruction accuracy, the latent representations of testing instances are enforced to be close to that of their neighbours from the training set via posterior regularization. Experiments on three fMRI recording datasets demonstrate that our approach can more accurately reconstruct visual stimuli.
分类： 计算机科学 >> 计算机应用技术 提交时间： 2016-05-03
摘要： 近年来，迁移学习已经引起了广泛的关注。迁移学习是运用已存有的知识对不同但相关领域问题进行求解的新的一种机器学习方法。传统机器学习基于两个基本假设：(1) 用于学习的训练样本与新的测试样本满足独立同分布的条件；(2) 必须有足够可利用的训练样本才能学习得到一个好的分类模型。迁移学习降低了要求，目的是迁移已有的知识来解决目标领域中仅有少量或没有有标签样本数据时的学习问题。本文对迁移学习算法以及相关理论研究进展进行了综述，并介绍了我们在该领域所做的研究工作，特别是利用生成模型在概念层面建立迁移学习模型。最后指出了迁移学习下一步可能的研究方向。
分类： 计算机科学 >> 计算机应用技术 提交时间： 2016-05-03
摘要： 深度学习是机器学习领域的一个新的研究方向，其核心思想在于模拟人脑的层级抽象结构，通过无监 督的方式从大规模数据（例如图像、声音和文本）中学习特征。近年来，深度学习在计算机视觉、语音识 别等研究领域取得的巨大成功使得研究者们对其寄予更多的关注。本文从深度学习的概念、发展历程、模 型、训练方法以及应用等几个方面对其进行概述，并对深度学习的未来发展做出展望。
分类： 计算机科学 >> 自然语言理解与机器翻译 分类： 图书馆学、情报学 >> 情报过程自动化的方法和设备 提交时间： 2019-10-29
摘要： Purpose: Move recognition in scientific abstracts is an NLP task of classifying sentences of the abstracts into different types of language unit. To improve the performance of move recognition in scientific abstracts, a novel model of move recognition is proposed that outperforms BERT-Base method. Design: Prevalent models based on BERT for sentence classification often classify sentences without considering the context of the sentences. In this paper, inspired by the BERT's Masked Language Model (MLM), we propose a novel model called Masked Sentence Model that integrates the content and contextual information of the sentences in move recognition. Experiments are conducted on the benchmark dataset PubMed 20K RCT in three steps. And then compare our model with HSLN-RNN, BERT-Base and SciBERT using the same dataset. Findings: Compared with BERT-Base and SciBERT model, the F1 score of our model outperforms them by 4.96% and 4.34% respectively, which shows the feasibility and effectiveness of the novel model and the result of our model comes closest to the state-of-the-art results of HSLN-RNN at present. Research Limitations: The sequential features of move labels are not considered, which might be one of the reasons why HSLN-RNN has better performance. And our model is restricted to dealing with bio-medical English literature because we use dataset from PubMed which is a typical bio-medical database to fine-tune our model. Practical implications: The proposed model is better and simpler in identifying move structure in scientific abstracts, and is worthy for text classification experiments to capture contextual features of sentences. Originality: The study proposes a Masked Sentence Model based on BERT which takes account of the contextual features of the sentences in abstracts in a new way. And the performance of this classification model is significantly improved by rebuilding the input layer without changing the structure of neural networks.
分类： 计算机科学 >> 计算机应用技术 提交时间： 2020-09-30
摘要： This paper puts forward the concept of Small Private Online Judge (SPOJ). Compared with Massive Open Online Judge (MOOJ), SPOJ has advantages in structured data acquisition of students' virtual behavior for its specific function and tight coupling with the classroom. SPOJ-based empirical education research can be conducted within "Acquisition-Analysis-Application" (3A) Framework. The case study of a SPOJ program clarifies the standard pattern of SPOJ-based 3A research and highlights the emergence of education-intelligence concept. The challenges of SPOJ-based empirical education research and implications of SPOJ are also discussed.
分类： 计算机科学 >> 计算机应用技术 提交时间： 2020-02-15
摘要： [目的]探究人工智能在新型冠状病毒（2019-nCoV）的诊断、治疗和控制中的应用场景和进展，以利用人工智能为新型冠状病毒肺炎的防控提供助力。 [方法]剖析新型冠状病毒肺炎防控的技术需求，从人工智能基因测序、辅助诊断、远程专家系统、药物筛查与研制等方面，分析当前的应用进展，挖掘应用的机遇。 [结果]中国是新型冠状病毒疫情最严重的国家，存在诸多的技术短板有待科技助力，AI能在疫情防控中发挥出重要的作用，但目前处于初步阶段，缺乏经过验证的落地成果；AI辅助诊断领域重复性研发较多，其他方面研究较少。 [局限]当前应用的数据大部分来自网站报道，如有更多的学术性成果，进展的分析将更全面。 [结论]需要加大投入和调控，在数据、算法和算力共享的基础上，各方面全面展开研发。
分类： 计算机科学 >> 计算机应用技术 提交时间： 2019-05-05
摘要： 机器人具有极大降低劳动体力需求的潜力，是目前研究热点之一。机器人设计是智力劳动密集型工业，因此机器人的自动设计方法研究具有重要的现实意义。由于现有研究中鲜有自动设计的研究，本文探讨自动设计的基本框架。将机器人设计抽象为黑盒参数优化的过程，通过物理模拟对一个特定的机器人设计进行评分，确定此次设计完成给定任务的能力。通过更改设计参数，进行完成任务能力的优化。即自动设计方法的输入是一组零件，一个需要机器人完成的任务以及一个一般是基于强化学习的机器人控制训练方法，输出是这一组零件的组合方式。由于物理模拟，或真实的实验验证在时间和劳动成本上较为昂贵，因此黑盒优化的过程须尽可能降低采样次数。本文采用贝叶斯参数优化完成设计任务。实验表明，所提出的方法可以在5次左右的模拟中完成一个两关节机器人自动设计任务。
分类： 计算机科学 >> 计算机应用技术 提交时间： 2019-11-23
摘要： Controlling weeds with reduced reliance on herbicides is one of the main challenges to move toward a more sustainable agriculture. Robotic weeding is a thought to be a viable way to reduce the environmental loading of agrochemicals while keeping the operation efficiency high. One of the key technologies for performing robotic weeding is automatic detection of crops and weeds in fields. This paper presents an overview on various methods for detecting plants based on machine vision, mainly concentrating on two main challenges: dealing with changing light and crop/weed discrimination. To overcome the first challenge, both physical and algorithmic methods have been proposed. Physical methods can result in a more cumbersome machine while algorithmic methods are less robust. For crop/weed discrimination, deep-learning-based methods have shown obvious advantages over traditional methods based on hand-crafted features. However, traditional methods still hold some merits that can be leveraged to deep-learning-based methods. With the fast development of hardware technologies, researchers should take full advantage of advanced hardware to ease the algorithm design. In the future, the identification of crops and weeds can be more accurate and fine-grained with the support of online databases and computing resources based on the advances in artificial intelligence and communication technologies.
分类： 计算机科学 >> 计算机应用技术 提交时间： 2020-02-24
zhidong,Xue
Lei,Zhao
Tailang,Yin
Yan,Fu
Zehua,Lyu
yiping,Dang
Yujiang,Zeng
Silou,Huang
Bing,Qu
Hongya,Lyu
Chen,Huang
Zhiyou,Kong
Kepei,Xu
Feipeng,Zhou
Hexun,Dong
He,Hu
Jing,Tang
Senyuan,Xue
Zhixiang,Fang
Jinxiang,Lu
摘要： Background：The COVID-19 Epidemic emerged in Wuhan, Hubei province, China. Ever since Wuhan lockdown on January 23rd, mass quarantines were exercised on Wuhan and other epidemic areas of China. We aimed to clarify how ordinary Wuhan people defend against COVID-19 epidemic at home through the Internet survey. Methods：A questionnaire survey, consisting of 30 questions were posted on the Internet. The following aspects were investigated: household preventive measures, self-monitoring of discomfort symptoms, immunity boosting against the epidemic, frequency and reasons of outgoing and mental status of the isolated people. The questionnaire was circulated on Wechat. We marked the areas based on the surveyed network IP addresses and categorized respondents into group A(Wuhan), B(Hubei Province excluding Wuhan ), C, and D based on the epidemic severity of their areas announced by Baidu.com at 17:00 on February 8, 2020. And a comparative study was conducted to illustrate how Wuhan people took the anti-COVID-19 strategies and how efficient these preventive measures were. Findings：In terms of discomfort symptoms, Wuhan, as Group A, had the lowest asymptomatic percentages (70.2%), compared to the average 78.5% (±7%). Considering the three typical symptoms for the COVID-19, i.e., cough, fever and fatigue, Wuhan (9.67%) greatly deviated from the average (7.68%). The fatigue was the most significant factor in the deviation, exceeding the average by 1.35%. In terms of household protection measures, most people or families were able to take effective protection measures with very low frequency of going out, but the percentage of those who took this practice was obviously smaller in Wuhan and Hubei Province. From the aspect of going out, most of the people in Wuhan only went out for shopping and work, with a small number of people for social gathering. In terms of immunity boosting, compared with Group C and D, it was relatively lower in Wuhan. Overall, most people chose to enhance their immunity through regular schedule, exercise, sufficient nutrition. Only 33.44% of people in Group A did not go out, and 59.97% had to go out for living supplies, which was the highest level among the four groups. However, the percentage of people who went out for work and unnecessary activities remains the lowest while 1% of the population went out for public welfare activities, higher than other groups. Worry about the family health topped all the parameters for all the groups. Among them, Wuhan has reached a maximum of 49.61%, higher than the average level of 36.62% (± 10.69%). Mental status except for feeling bored and lonely were the highest in Wuhan. Suggestions：When the epidemic prevention and control is still in a sticky state, and Wuhan started a stricter control measure, the closed management of communities, on Feb 11, 2020, it is expected that our findings can provide some insights into the current household preventive actions and arouse more attentions of the public to some ignored preventive precautions. Unnecessary outgoing should be strictly abandoned. Regular schedule, exercises and nutrition were the top 3 measures participants would choose to enhance their own immunity system. It seems that people in Wuhan would choose nutrition and regular scheduler rather than exercises as the primary immunity-boosting ways. Exercise should be especially advocated as an effective way to enhance the immunity system. In terms of physical condition, people in Wuhan should take more active measures when symptoms occurred. The mentality is also an important aspect requiring intensive attention with the conduct of stricter control management in Wuhan while the rest groups gradually resume to work and ordinary life.
分类： 计算机科学 >> 自然语言理解与机器翻译 提交时间： 2019-05-12
摘要： Abstract. Computational chemistry develops fast in recent years due to the rapid growth and breakthroughs in AI. Thanks for the progress in natural language processing, researchers can extract more fine-grained knowledge in publications to stimulate the development in computational chemistry. While the works and corpora in chemical entity extraction have been restricted in the biomedicine or life science field instead of the chemistry field, we build a new corpus in chemical bond field anno- tated for 7 types of entities: compound, solvent, method, bond, reaction, pKa and pKa value. This paper presents a novel BERT-CRF model to build scientific chemical data chains by extracting 7 chemical entities and relations from publications. And we propose a joint model to ex- tract the entities and relations simultaneously. Experimental results on our Chemical Special Corpus demonstrate that we achieve state-of-art and competitive NER performance.
分类： 心理学 >> 应用心理学 分类： 计算机科学 >> 计算机应用技术 提交时间： 2020-03-08
摘要： 人格影响着个体的工作生活方式，对于个体的心理疏导、职业发展等具有重要指导意义。传统方法通过量表测评人格得分存在个体拒绝回答、盲目作答等问题，近年来随着机器学习的发展为人格识别提供了新的思路。本文使用被试者自我介绍视频和大五人格量表得分，经过关键点提取、特征降维、建模、迭代调参等步骤，针对不同人格维度得到不同的预测模型。测试结果表明，基于自我介绍视频的人格预测模型在各维度都接近或达到中等相关，能够提供无侵扰的人格自动识别，为人格测量提供了新的思路。
分类： 心理学 >> 应用心理学 分类： 计算机科学 >> 计算机应用技术 提交时间： 2019-12-20
摘要： [背景]LIWC（基于语词计量的文本分析）以关键词的词频统计为基础，可对个体和群体的表达语句的心理学意义等方面进行量化分析。由于文言文的表达方式与现代汉语存在明显的差异，为了分析文言文文本的心理学意义，我们在简体中文LIWC词典（Simplified Chinese LIWC 2015年版本, 简称SC-LIWC）的基础上，构建了古文LIWC（Classical Chinese LIWC，以下简称CC-LIWC）词典。[目的]本研究的目的是探究如何构建CC-LIWC词典并介绍如何使用该词典对古文文本进行分析。[方法]获取在线汉语词典的全部词汇及其对应解释，保留文言文词及其现代文译文，并从译文中寻找SC-LIWC词，将SC-LIWC词与文言文词进行匹配。对匹配结果进行人工标注，确保结果的一致性与准确性。[结果]最终生成的CC-LIWC包含了81个词类与49136个文言文词条。[局限]古文中一词多义、一词多性的情况较为普遍，对词典中词汇的分类存在一定影响。[结论]使用CC-LIWC对《论语（节选）》、《孤愤》进行词频分析，分析结果体现了儒家的中庸与法家的注重逻辑辩证的区别，说明CC-LIWC词典能够有效区分文本的表达倾向。
分类： 心理学 >> 应用心理学 分类： 计算机科学 >> 计算机应用技术 提交时间： 2020-01-07
摘要： 古文献的研究有助于传统文化的继承与发扬，而古文分词则是利用自然语言处理技术对古文献进行分析的重要环节，但由于缺少规范的数据资料而没有像现代汉语分词取得突破性进展。当前互联网拥有大量古汉语文本和词典方面的数据资料，但是这些数据分散，没有得到有效地整合。本文提出采集互联网非结构化古汉语数据，经过数据清洗和预处理抽取出一个古汉语基础词典，然后再利用互信息、信息熵、位置成词概率相结合的新词发现方法从大规模古籍文本中抽取古汉语候补词典，最终将基础词典与候补词典融合，利用正向最大匹配实现对古文的分词。与开源的分词器甲言在基于词典的分词方面比较后F值提高了14%，取得了良好的效果，结果证明本文构建的分词器可以应用在古汉语文本分词上。
分类： 计算机科学 >> 计算机应用技术 提交时间： 2017-03-10
摘要： 本报告回顾中国高性能计算机系统研发和应用的历史，介绍目前政府所支持的高性能计算项目、高性 能计算中心、主要研究机构、重要的高性能计算应用领域和国内厂家的现状。从系统研制、应用开发和长期规划方面，与美国、欧洲和日本发达国家的高性能计算进行了比较。此外，本报告还对中国在高性能计 算方面的技术和应用发展方向进行了预测。
分类： 心理学 >> 社会心理学 分类： 计算机科学 >> 计算机应用技术 提交时间： 2018-03-15
摘要： 理清“一带一路”沿线国家或地区的民心特点，并找到有效的合作交往模式，是关系到国家战略实施的重大问题。但是，由于地域辽阔、民族众多，且地缘政治、经济、文化因素（如原苏联影响、欧美国家殖民、宗教传统等）异常复杂，传统的分析方法往往难以奏效。该研究结合文化心理学和大数据分析技术，利用社交媒体Twitter数据来分析“一带一路”沿线国家或地区的自我表征特点（独立性或个人主义），并建立自我表征与社会信任（普遍信任、特殊信任）的预测模型，以探究与“一带一路”沿线国家或地区合作交往的行为模式，即：自我表征是独立，还是互依；人际关系偏好是陌生人之间的普遍信任，还是熟人间的特殊信任。结果表明，“一带一路”沿线国家或地区在自我独立性这一个人主义文化指标上存在较大的变异，且主要受欧美国家殖民历史和当地宗教传统的影响；此外，针对陌生人、外国人的普遍信任与针对家人、熟人的特殊信任，可以通过个人主义指标来预测。总之，“一带一路”沿线的文化是多样的，可以通过社交媒体产生的海量语料库快速计算其个人主义指标，并以此来建立自我表征与社会信任的预测模型。该研究为分析“一带一路”战略区域的“民心”特点、探索当地合作交往的行为模式提供了新的技术路径。
分类： 计算机科学 >> 计算机应用技术 提交时间： 2020-04-14
摘要： Low-light images suffer from severe noise and low illumination. Current deep learning models that are trained with real-world images have excellent noise reduction, but a ratio parameter must be chosen manually to complete the enhancement pipeline. In this work, we propose an adaptive low-light raw image enhancement network to avoid parameter-handcrafting and to improve image quality. The proposed method can be divided into two sub-models: Brightness Prediction (BP) and Exposure Shifting (ES). The former is designed to control the brightness of the resulting image by estimating a guideline exposure time t 1 . The latter learns to approximate an exposure-shifting operator ES, converting a low-light image with real exposure time t 0 to a noise-free image with guideline exposure time t 1 . Additionally, structural similarity (SSIM) loss and Image Enhancement Vector (IEV) are introduced to promote image quality, and a new Campus Image Dataset (CID) is proposed to overcome the limitations of the existing datasets and to supervise the training of the proposed model. In quantitative tests, it is shown that the proposed method has the lowest Noise Level Estimation (NLE) score compared with BM3D-based low-light algorithms, suggesting a superior denoising performance. Furthermore, those tests illustrate that the proposed method is able to adaptively control the global image brightness according to the content of the image scene. Lastly, the potential application in video processing is briefly discussed.
分类： 物理学 >> 电磁学、光学、声学、传热、经典力学和流体动力学 分类： 计算机科学 >> 信息安全 提交时间： 2017-03-28
摘要： Classical optical hiding methods are symmetric, being apt to realize but not secure. The security is improved in existing non-symmetric hiding techniques, yet all of them fails in convenient extractions, still not optically realized so far. Here, we propose an asymmetric optical hiding method based on visual cryptography, achieving the high security and the easy extraction at the same time. In the hiding process, we convert the secret information into a set of fabricated phase-keys, which are completely independent of each other, intensity-detected-proof, and image-covered, this complex hiding procedure leading to the high security. Correspondingly, during the extraction process, the covered phase-keys are illuminated with laser beams and then incoherent superposed to extract the hidden information directly by human visual system, without complicated optical implementations and any additional computation, resulting in the convenience of extracting. Optical experiments verify that both the high security and the easy extraction are obtainable in the visual-cryptography-based optical hiding.
分类： 计算机科学 >> 计算机科学的集成理论 提交时间： 2019-04-01 合作期刊: 《计算机应用研究》
摘要： 针对蚁群融合模糊C-means (FCM)聚类算法在蛋白质相互作用网络中进行复合物识别的准确率不高、召回率较低以及时间性能不佳等问题进行了研究，提出一种基于模糊蚁群的加权蛋白质复合物识别算法FAC-PC (algorithm for identifying weighted protein complexes based on fuzzy ant colony clustering)。首先，融合边聚集系数与基因共表达的皮尔逊相关系数构建加权网络；其次提出EPS (essential protein selection)度量公式来选取关键蛋白质，遍历关键蛋白质的邻居节点，设计蛋白质适应度PFC (protein fitness calculation)来获取关键组蛋白质，利用关键组蛋白质替换种子节点进行蚁群聚类，克服蚁群算法中因大量拾起放下和重复合并过滤操作而导致准确率和收敛速度过慢的缺陷；接着设计相似度SI (similarity improvement)度量优化拾起放下概率来对节点进行蚁群聚类进而获得聚类数目；最后将关键蛋白质和通过蚁群聚类得到的聚类数目初始化FCM算法，设计隶属度更新策略来优化隶属度的更新，同时提出兼顾类内距和类间距的FCM迭代目标函数，最终利用改进的FCM完成复合物的识别。将FAC-PC算法应用在DIP数据上进行复合物的识别，实验结果表明FAC-PC算法的准确率和召回率较高，能够较准确地识别蛋白质复合物。
分类： 计算机科学 >> 计算机应用技术 提交时间： 2020-02-03
摘要： 自从武汉市爆发新型冠状病毒疫情以来，迅速蔓延的事态已经造成300多人死亡，一万多人感染。在中国之外有一百多起病例，影响了全球十几个国家。研究人员已经报道了冠状病毒的全基因组序列，并且正在迅速开发快速诊断试剂盒、有效的治疗方法以及预防性疫苗。最初快速增长的确诊病例触发了武汉及附近城市的封锁。世界各地的科学家尝试建立数学模型来预测未来几天内的感染病例数。但是，交通和文化习俗等主要因素尚未得到足够的权衡。我们的模型并不是为了精确预测感染病例数量，而是旨在模拟公共流行紧急情况下的动态情况以及不同影响因素的贡献。我们希望我们的模型和模拟能够为全球公共卫生机构提供更多的见解和观点信息，以便设计出更好的预防和控制解决方案。
点击量
待评议
点击量
下载量
评论
通过
