Neural and Evolutionary ComputingMulti-rendezvous Spacecraft Trajectory Optimization with Beam P-ACOLuís F. Simões,Dario Izzo,Evert Haasdijk,A. E. EibenComments: Code available atthis https URLSubjects:Neural and Evolutionary Computing (cs.NE); Space Physics (physics.space-ph)The design of spacecraft trajectories for missions visiting multiplecelestial bodies is here framed as a multi-objective bilevel optimizationproblem. A comparative study is performed to assess the performance ofdifferent Beam Search algorithms at tackling the combinatorial problem offinding the ideal sequence of ...继续阅读 (14)

1. 我爱机器学习(1)arXiv Paper Daily: Tue, 4 Apr 20172. ALGOBEANS(1)Time Series Analysis with Generalized Additive Models3. 新智元(4)DeepMind 的下一场博弈：用机器学习颠覆资产管理？【干货】深度学习必备：随机梯度下降（SGD）优化算法及可视化【无人车排行】谷歌、Uber、特斯拉没进Top 5，百度名列“挑战者”【深度】机器的崛起：算法会超出我们的控制吗？4. 机器之心(4)教程 | 新手指南：如何在AWS GPU上运行Jupyter noterbook？从强化学习基本概念到Q学习的实现，打造自己的迷宫智能体解读 | 谷歌像素递归超分辨率研究：怎么消灭低分辨率图像马赛克？干货 | 物体检测算法全概述：从传统检测方法到深度神经网络框架5. 大数据文摘(3)苹果自研芯片传言证实，被“抛弃”的Imagination股价跌70%视频更新｜斯坦福CS231n深度学习与计算机视觉课时31-来自Jeff Dean的受邀报告（下）中国信通院发布《中国大数据发展调查报告（2017年）》| 附下载6. 算法与数学之美(2)小波变换支持向量机（三）核函数 ...继续阅读 (14)

Neural and Evolutionary ComputingOn Self-Adaptive Mutation Restarts for Evolutionary Robotics with Real RotorcraftGerard David HowardComments: 8 pagesSubjects:Neural and Evolutionary Computing (cs.NE); Robotics (cs.RO)Self-adaptive parameters are increasingly used in the field of EvolutionaryRobotics, as they allow key evolutionary rates to vary autonomously in acontext-sensitive manner throughout the optimisation process. A significantlimitation to self-adaptive mutation is that rates can be set unfavourably,which hinders convergence. Rate restarts are typically employed to remedy this,but t ...继续阅读 (25)

1. 新智元(3)【干货】神经机器翻译全流程解析，one-shot 和 zero-shot 学习成亮点吴恩达错了！实践者说，企业千万别设“首席 AI 官”深圳“AI”领袖峰会：马云称机器智能应成为人类合伙人，李彦宏、马化腾和沈向洋等聚焦AI2. 机器之心(5)机器理解大数据的秘密：聚类算法深度详解PaperWeekly 第31期 | 远程监督在关系抽取中的应用MIT与波士顿大学联合报告：事实证明机器人确实会降低就业和工资学界 | 阿里人工智能新研究：在星际争霸中实现多兵种协同作战专栏 | 超越图灵测试：自省性问答才是检测真正人工智能的方法3. 大数据文摘(2)BAT三巨头激辩AI：李彦宏称强人工智能不会到来；马云说So TM WhatTED演讲 | 数据滥用时代，3招教你辨别身边不靠谱数据4. 算法与数学之美(2)支持向量机SVM（二）Math Horizons的愚人节笑话：一个奇妙的分式方程 ...继续阅读 (17)

1、【百度发布DuerOS智慧芯片，AI产业化方向靠算法+芯片】百度联合ARM、紫光展锐和汉枫电子发布DuerOS智慧芯片，这是百度度秘事业部今年独立以来，发布的首个重要产品。DuerOS智慧芯片拥有低成本芯片和模组，将自带DuerOS，可以放在任何硬件中。<点评：百度用“算法+芯片”的组合实现人工智能产业化落地>相关链接：百度DuerOS智慧芯片发布 可对话“智慧设备”时代来临2、【小程序将上线附近的小程序：用完即走还要让用户再来】微信团队宣布小程序将上线两项全新能力，“第三方平台”和“附近的小程序”。“第三方平台”开放后，可以让商户的小程序开发更省心，也方便第三方批量管理商户。通过“附近的小程序”，用户可以在微信中快速找到附近的小程序和服务，同时也能够帮助线下商户更直接地触达用户，让小程序融入更多生活场景。<点评：把之前一直受开发者和用户诟病的几个“过分克制”的地方开放了，而且开放的力度还不小>相关链接：小程序将上线附近的小程序：用完即走还要让用户再来3、【全面屏、无Home键、无线充、面部识别，三星S8留给苹果的设计不多了】Galaxy S8/S8+采用5.8/6.2英寸18.5:9圆角2K双曲面屏。4+64GB/6+128GB，800W+1200W主相机，3000mAH/3500mAH，无线快充。IP68、虹膜识别，背面指纹，虚拟压感按键。<点评：三星S8留给苹果的设计不多 ...继续阅读 (18)

Neural and Evolutionary ComputingApplication of a Shallow Neural Network to Short-Term Stock TradingAbhinav Madahar,Yuze Ma,Kunal PatelComments: 4 pagesSubjects:Neural and Evolutionary Computing (cs.NE)Machine learning is increasingly prevalent in stock market trading. Thoughneural networks have seen success in computer vision and natural languageprocessing, they have not been as useful in stock market trading. Todemonstrate the applicability of a neural network in stock trading, we made asingle-layer neural network that recommends buying or selling shares of a stockby comparing the highest hi ...继续阅读 (20)

Neural and Evolutionary ComputingHierarchical Surrogate Modeling for Illumination AlgorithmsAlexander HaggSubjects:Neural and Evolutionary Computing (cs.NE)Evolutionary illumination is a recent technique that allows producing manydiverse, optimal solutions in a map of manually defined features. To supportthe large amount of objective function evaluations, surrogate model assistancewas recently introduced. Illumination models need to represent many more,diverse optimal regions than classical surrogate models. In this PhD thesis, wepropose to decompose the sample set, decreasing model complexity ...继续阅读 (26)

Neural and Evolutionary ComputingExperimental Analysis of Design Elements of Scalarizing Functions-based Multiobjective Evolutionary AlgorithmsMansoureh Aghabeig,Andrzej JaszkiewiczSubjects:Neural and Evolutionary Computing (cs.NE)In this paper we systematically study the importance, i.e., the influence onperformance, of the main design elements that differentiate scalarizingfunctions-based multiobjective evolutionary algorithms (MOEAs). This class ofMOEAs includes Multiobjecitve Genetic Local Search (MOGLS) and MultiobjectiveEvolutionary Algorithm Based on Decomposition (MOEA/D) and proved t ...继续阅读 (28)

Neural and Evolutionary ComputingWhere to put the Image in an Image Caption GeneratorMarc Tanti(1),Albert Gatt(1),Kenneth P. Camilleri(1) ((1) University of Malta)Comments: under review, 29 pages, 5 figures, 6 tablesSubjects:Neural and Evolutionary Computing (cs.NE); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)When a neural language model is used for caption generation, the imageinformation can be fed to the neural network either by directly incorporatingit in a recurrent neural network — conditioning the language model byinjecting image features R ...继续阅读 (23)

Neural and Evolutionary ComputingPonyGE2: Grammatical Evolution in PythonMichael Fenton,James McDermott,David Fagan,Stefan Forstenlechner,Michael O'Neill,Erik HembergComments: 8 pages, 4 figures, submitted to the 2017 GECCO Workshop on Evolutionary Computation Software Systems (EvoSoft)Subjects:Neural and Evolutionary Computing (cs.NE)Grammatical Evolution (GE) is a population-based evolutionary algorithm,where a formal grammar is used in the genotype to phenotype mapping process.PonyGE2 is an open source implementation of GE in Python, developed at UCD’sNatural Computing Research a ...继续阅读 (23)

1. 机器之心(4)一周论文 | 解读2016年最值得读的三篇NLP论文 + 在线Chat实录机器之心「GMIS 2017」嘉宾揭秘：LSTM之父Jürgen Schmidhuber业界 | 自动驾驶事故频发：Uber步特斯拉、谷歌之后尘综述 | 一文帮你发现各种出色的GAN变体 ...继续阅读 (11)

1. 新智元(5)DeepMind 新算法NEC：让 AI 在学习速度上超越人类特朗普政府漠视 AI 浪潮，中国能否抓住时机？PaddlePaddle发布基于Docker的AI系统开发流程BAT 齐聚智能+中国主平台，新智元&英特尔 AI 技术峰会参会指南【最详尽的GAN介绍】王飞跃等：生成式对抗网络 GAN 的研究进展与展望2. 机器之心(5)OpenAI详解进化策略方法：可替代强化学习深度 | David Silver全面解读深度强化学习：从基础概念到AlphaGo业界 | 深度学习芯片公司Graphcore初探：颠覆GPU、FPGA和CPU的新一代处理器IPU教程 | 如何使用TensorFlow构建、训练和改进循环神经网络业界 | 百度计划新增第二个硅谷人工智能研发机构，张亚勤全面负责领导3. 算法与数学之美(3)梅森素数：数学海洋中的璀璨明珠泛函分析的名字怎么来的？清华教授、肌肉男神、88篇论文被SCI收录…4. 算法与数学之美(3)梅森素数：数学海洋中的璀璨明珠泛函分析的名字怎么来的？清华教授、肌肉男神、88篇论文被SCI收录… ...继续阅读 (17)

Neural and Evolutionary ComputingClassification-based RNN machine translation using GRUsRi Wang,Maysum Panju,Mahmood GohariComments: 7 pages, 1 figure; graduate course research projectSubjects:Neural and Evolutionary Computing (cs.NE); Learning (cs.LG)We report the results of our classification-based machine translation model,built upon the framework of a recurrent neural network using gated recurrentunits. Unlike other RNN models that attempt to maximize the overall conditionallog probability of sentences against sentences, our model focuses aclassification approach of estimating the conditio ...继续阅读 (20)

Neural and Evolutionary ComputingASP: Learning to Forget with Adaptive Synaptic Plasticity in Spiking Neural NetworksPriyadarshini Panda,Jason M. Allred,Shriram Ramanathan,Kaushik RoyComments: 14 pages, 14 figuresSubjects:Neural and Evolutionary Computing (cs.NE); Computer Vision and Pattern Recognition (cs.CV)A fundamental feature of learning in animals is the “ability to forget” thatallows an organism to perceive, model and make decisions from disparate streamsof information and adapt to changing environments. Against this backdrop, wepresent a novel unsupervised learning mechan ...继续阅读 (24)

1. 我爱机器学习(1)arXiv Paper Daily: Thu, 23 Mar 20172. 新智元(5)Facebook何凯明最新研究：通用对象分割框架Mask R-CNN，COCO三项最佳FPGA 超越 GPU，问鼎下一代深度学习主引擎【重磅】腾讯任命张潼为AI Lab主任，“无”恩达时代BAT人工智能三国杀（投票）投资专家要小心，机器学习正在入侵VCYann LeCun清华演讲：生成对抗学习最新进展与如何让机器有常识（90 PPT 下载）3. 机器之心(5)邀请函 | 2017国家机器人发展论坛演讲 | Yann LeCun清华演讲：深度学习与人工智能的未来机器之心专访腾讯「绝艺」团队负责人：用全新强化学习方法造就更强模型业界 | 腾讯AI Lab启动首届学术论坛，正式宣布张潼出任实验室主任开发者调研 | Rust最受欢迎、Python最受关注、机器学习专家收入最高4. 大数据文摘(1)昨天没能去清华亲眼见到Yann LeCun？这有一份演讲精华（提纲+PPT）5. 算法与数学之美(3)数学专业经典读物不能更简单通俗的机器学习基础名词解释中国互联网的二十年（深度好文） ...继续阅读 (25)

Neural and Evolutionary ComputingAn Accelerated Analog Neuromorphic Hardware System Emulating NMDA- and Calcium-Based Non-Linear DendritesJohannes Schemmel,Laura Kriener,Paul Müller,Karlheinz MeierComments: Accepted at IJCNN 2017Subjects:Neural and Evolutionary Computing (cs.NE); Emerging Technologies (cs.ET)This paper presents an extension of the BrainScaleS accelerated analogneuromorphic hardware model. The scalable neuromorphic architecture is extendedby the support for multi-compartment models and non-linear dendrites. Thesefeatures are part of a SI{65}{anometer} prototype ASIC. It a ...继续阅读 (22)

Neural and Evolutionary ComputingEmpirical Analysis of the Necessary and Sufficient Conditions of the Echo State PropertySebastián BasterrechComments: 23 pages, 14 figures, accepted paper for the IEEE IJCNN, 2017Subjects:Neural and Evolutionary Computing (cs.NE)The Echo State Network (ESN) is a specific recurrent network, which hasgained popularity during the last years. The model has a recurrent networknamed reservoir, that is fixed during the learning process. The reservoir isused for transforming the input space in a larger space. A fundamental propertythat provokes an impact on the m ...继续阅读 (38)

Neural and Evolutionary ComputingImplicit Gradient Neural Networks with a Positive-Definite Mass Matrix for Online Linear Equations SolvingKe ChenComments: Submitted to Information Processing LettersSubjects:Neural and Evolutionary Computing (cs.NE); Systems and Control (cs.SY)Motivated by the advantages achieved by implicit analogue net for solvingonline linear equations, a novel implicit neural model is designed based onconventional explicit gradient neural networks in this letter by introducing apositive-definite mass matrix. In addition to taking the advantages of theimplicit neural dynam ...继续阅读 (31)

1. 我爱机器学习(1)Machine Learning Daily: 2017-03-192. 新智元(5)杜克大学陈怡然教授：关于AI人才八位数年薪的几句话谷歌开源JPEG压缩算法，比同类方法性能优越35%【重磅】腾讯“绝艺”9连胜横扫日韩AI，距世界冠军一步之遥【资源】深度学习论文阅读路线图，从经典到前沿尽览【荐书】机器学习模型评估：关键概念和可能碰到的陷阱指南（附下载）3. 机器之心(5)PaperWeekly 第29期 | 你的Emoji不一定是我的EmojiAIWTB 2017开发者大会前瞻：线上门票免费送从重采样到数据合成：如何处理机器学习中的不平衡分类问题？学界 | MIT新型多机器人系统安全技术：可保护自动驾驶汽车和无人机等免受通信攻击资源 | TensorFlow极简教程：创建、保存和恢复机器学习模型4. 大数据文摘(3)TED演讲 | 我知道统计学很重要，但是关我什么事？免费限时 | 2017云栖大会深圳峰会报名中 揭秘云产业奥斯卡德国机器人0.637秒破解魔方打破世界纪录5. 算法与数学之美(3)21张GIF动图让你秒懂数学原理目标跟踪算法日本81岁老太退休后自学编程 成功开发一款游戏6. 算法与数学之美(3)21张GIF动图让你秒懂数学原理目标跟踪算法日本81岁老太退休后自学编程 成功开发一款游戏 ...继续阅读 (16)

1. 我爱机器学习(1)Machine Learning Daily: 2017-03-182. 大数据文摘(4)交通数据分析帮你避开99%的酒驾司机专属赠票 | 2017第九届云计算中国峰会暨混合云世界论坛倒计时人工智能概念第一股？即将在美国上市的Veritone是怎样一家公司特写｜Kaggle被谷歌收购能否复制，中国的Kaggle们走到哪了3. 算法与数学之美(3)薛定谔的滚与深度学习中的物理AI工程师年薪百万？大数据告诉你真相女朋友掉水里，各类程序猿怎么救？4. 算法与数学之美(3)薛定谔的滚与深度学习中的物理AI工程师年薪百万？大数据告诉你真相女朋友掉水里，各类程序猿怎么救？5. 布洛卡区(1)NeuralFinder:集成人工生命和遗传算法自动发现神经网络最优结构 ...继续阅读 (17)

Neural and Evolutionary ComputingLarge Scale Evolution of Convolutional Neural Networks Using Volunteer ComputingTravis DesellComments: 17 pages, 13 figures. Submitted to the 2017 Genetic and Evolutionary Computation Conference (GECCO 2017)Subjects:Neural and Evolutionary Computing (cs.NE)This work presents a new algorithm called evolutionary exploration ofaugmenting convolutional topologies (EXACT), which is capable of evolving thestructure of convolutional neural networks (CNNs). EXACT is in part modeledafter the neuroevolution of augmenting topologies (NEAT) algorithm, withnotable exceptio ...继续阅读 (25)

1. 我爱机器学习(2)Machine Learning Daily: 2017-03-17arXiv Paper Daily: Fri, 17 Mar 2017 ...继续阅读 (12)

Neural and Evolutionary ComputingDeep learning with convolutional neural networks for brain mapping and decoding of movement-related information from the human EEGRobin Tibor Schirrmeister,Jost Tobias Springenberg,Lukas Dominique Josef Fiederer,Martin Glasstetter,Katharina Eggensperger,Michael Tangermann,Frank Hutter,Wolfram Burgard,Tonio BallSubjects:Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)Deep learning with convolutional neural networks (deep ConvNets) hasrevolutionized computer vision through end-to-end learning, i.e. learning fromthe raw data. Now, there is increasing i ...继续阅读 (28)

1. 我爱机器学习(2)arXiv Paper Daily: Thu, 16 Mar 2017Machine Learning Daily: 2017-03-16 ...继续阅读 (24)

1. 我爱机器学习(1)Machine Learning Daily: 2017-03-152. 深度学习世界(2)Keras 2发布：实现与TensorFlow的直接整合为机器赋予记忆：DeepMind重磅研究提出弹性权重巩固算法 ...继续阅读 (12)

Neural and Evolutionary ComputingComparison of echo state network output layer classification methods on noisy dataAshley PraterComments: 8 pages. International Joint Conference on Neural Networks (IJCNN 2017)Subjects:Neural and Evolutionary Computing (cs.NE)Echo state networks are a recently developed type of recurrent neural networkwhere the internal layer is fixed with random weights, and only the outputlayer is trained on specific data. Echo state networks are increasingly beingused to process spatiotemporal data in real-world settings, including speechrecognition, event detection, and ro ...继续阅读 (27)

1. 我爱机器学习(2)Machine Learning Daily: 2017-03-14arXiv Paper Daily: Tue, 14 Mar 20172. MachineLearningMastery(1)How to Install a Python 3 Environment on Mac OS X for Machine Learning and Deep Learning3. 深度学习世界(2)只需十四步：从零开始掌握Python机器学习（附资源）机器学习初创公司出路，投资人看好医疗诊断、计算生物和供应链等 9 大方向 ...继续阅读 (18)

Neural and Evolutionary ComputingEvolutionary Image Composition Using Feature Covariance MatricesAneta Neumann,Zygmunt L. Szpak,Wojciech Chojnacki,Frank NeumannSubjects:Neural and Evolutionary Computing (cs.NE)Evolutionary algorithms have recently been used to create a wide range ofartistic work. In this paper, we propose a new approach for the composition ofnew images from existing ones, that retain some salient features of theoriginal images. We introduce evolutionary algorithms that create new imagesbased on a fitness function that incorporates feature covariance matricesassociated with dif ...继续阅读 (22)

1. 我爱机器学习(2)arXiv Paper Daily: Mon, 13 Mar 2017Machine Learning Daily: 2017-03-132. 深度学习世界(2)【干货】ImageNet 历届冠军最新评析：哪个深度学习模型最适合你？（附资源下载）教程 | 如何转行成为一名数据科学家？ ...继续阅读 (23)

1. MachineLearningMastery(1)How to Setup a Python Environment for Machine Learning and Deep Learning with Anaconda ...继续阅读 (20)