HackerPapers

HN 二十年,最值得读的论文。

从 Hacker News 2006 年以来的全部投稿里,挑出链接本身就是论文的帖子:arXiv、期刊、会议论文、技术报告。每篇附中文标题,讨论最热的那些再加上中文导读和高赞评论翻译。

从必读开始
  • 8,000+篇被 HN 讨论过的论文

Essential

必读 30 篇

在论文库里看全部精选

编辑精选里 HN 讨论最热的 30 篇,每篇都有中文导读。不知道从哪读起,就从这里开始。

  1. 正在载入论文…

Archive

论文库

点柱子按年份筛选。注:源数据里 2023 年的帖子分数记录不全,2026 年尚未过完,这两年收录偏少。

难度
排序

    About

    这些论文是怎么挑出来的

    1. 01

      全量扫一遍

      读取 Hacker News 自 2006 年以来的全部 446 万条链接投稿,找出指向 arXiv、期刊、会议、技术报告和学术 PDF 的帖子。

    2. 02

      只留论文本身

      由 AI 逐条判断,剔除介绍论文的新闻、博客和幻灯片,同一篇论文的多次投稿合并计算,并统计它在评论区被引用的次数。

    3. 03

      写导读、译评论

      讨论最热的 1,500 篇配上中文导读,并精选、翻译 HN 高赞评论。导读由 AI 生成,可能有误,请以原文为准。

    经济与社会科学2014sciencedirect.com

    Tailspotting:识别并利用 CEO 度假行程获利

    Tailspotting: Identifying and profiting from CEO vacation trips

    入门 · 约 45 分钟读完前置知识:金融市场基础、统计基础

    读论文 HN 讨论
    232HN 得分
    54条评论

    导读

    这篇论文研究能否通过追踪 CEO 的度假出行信息来预测并从股价变动中获利,即利用公开的航班追踪数据识别高管行程并分析其与股票收益的关系。论文标题暗示这类'CEO 度假信号'可以带来可观的交易利润。

    HN 关注点:HN 读者对利用公开数据(如 ADS-B 飞行追踪)进行套利和情报分析非常感兴趣,讨论焦点集中在航班数据的技术可行性、类似技术的其他应用,以及对'发表即失效'的市场获利策略的质疑。

    关键词 CEO行程ADS-B航班追踪量化交易

    HN 评论精选

    这几条评论里:质疑 1 补充 2 亲历 1

    亲历亲历者介绍 ADS-B 飞行数据公开可接收,追踪公司高管航班推断股价完全可行

    几年前读到 ADS-B 在美国成为强制要求后,我立刻想到这一点,并花了几个月痴迷研究,成了我近年最有趣的爱好之一。ADS-B 本质上是飞机上的发射器,周期性报告注册号、高度、GPS 坐标和空速,数据实时且不像 FAA 数据源那样延迟或过滤。它不加密,任何有天线和设备的人都能接收通常 50-250 英里半径内所有装备飞机的位置。虽然许多美国国内飞机尚未装备,但到 2022 年必须全部装备;任何进入欧洲空域的飞机都必须装备,因此实际上所有国际航班都可追踪——包括大多数公务机。解码在软件层面非常简单,我运营着几个接收站……

    phil21查看英文原文

    This was what came to immediate mind after reading about ADS-B becoming mandatory in the US a few years back. This started a small few-month long obsession to understand it all, and became one of my more fun hobbies in the past few years. It already is mandatory in Europe. I didn't think of the vacationing correlation to stock prices, but simply tracking corporate movements and inferring future stock price movement from those movements. "Wall Street" of course coming to immediate mind here. For those not aware, ADS-B is essentially a transmitter in aircraft that periodically report registration number, altitude, GPS coordinates, and airspeed. This is to supplement the current air traffic control system, and eventually take it over. This data is in real time, and not delayed or censored like the FAA supplied data feeds are. It is not encrypted, so anyone with an antenna and the proper equipment can get the location information of every equipped aircraft within usually a 50-250mi radius, depending on antenna height, receiver quality, topology, etc. While many US domestic aircraft are not so equipped, this is rapidly changing as I believe all must be by 2022. Any aircraft that ever enters European airspace must be equipped, so that means for practical purposes any International flight can be tracked. This includes most corporate jets which are used to fly anywhere but within the United States. It's trivial to decode in software, and I run a couple installations with hacked up USB DBS TV dongles. They are not the highest quality (a few enterprising folks sell custom ASIC/boards for doing this better), but are cheap and easy to make work. It's a fun project to hack on for a weekend, I learned a bit about software defined radio among other things. Even fixed a bug in some code that accidentally calculated the aircraft's location to the wrong hemisphere :) Turns out GPS math is somewhat interesting, in the way they encode the data for the little bandwidth available. This is also the data that powers flightradar24, and probably some other sites out there. Basically a network of enthusiasts that run antenna installations reporting their data to a central server for visualization. I brought a small antenna and box on vacation with me to Jamaica a couple years back, and was able to track aircraft a good 300nm+ off the coast from the beach. It was fun watching the tracks route around storms and such.

    补充补充:同样的航班追踪技术还被用于调查军火走私和 CIA 秘密航班

    类似技术也可用于识别国际军火商、规避制裁者和 CIA 的引渡航班等。Alex Harrowell 偶尔会写这方面的博客,并附上几个链接:飞机调查资源、可疑的伊朗飞叙利亚航班、沙漠中神秘的 727 等。

    pjc50查看英文原文

    You can use a similar technique for identifying international arms dealers, sanction busters, CIA rendition flights, etc. Alex Harrowell occasionally blogs about this: http://www.harrowell.org.uk/blog/2011/01/09/a-quick-howto/ - aircraft investigation resources http://www.harrowell.org.uk/blog/2012/07/23/o-rly/ - Suspicious Iran-to-Syria flights http://www.harrowell.org.uk/blog/2009/12/23/mystery-jet-update-malian-727/ - mystery 727 in a desert

    质疑质疑:声称能从股市获利的论文必然无效,否则作者不会发表

    我从大学的统计/机器学习研究员那里听过一条最佳建议:如果一篇论文声称展示了在股市获利的技术,那它就是不管用的——真管用的话他们就不会发表了。话虽如此,这个领域有很多有趣的论文,从摘要看这篇的思路也挺有意思。

    pserwylo查看英文原文

    One of the best pieces of advice I heard from our resident stats/machine learning researcher at Uni was this: If a paper is published which claims to show a technique for profiting on the share market, then it doesn't work. If it did, they wouldn't have published it. Having said that, there is plenty of interesting papers in this area, and from the abstract, this paper seems to take an interesting approach too.

    补充补充案例:2008 年有人靠追踪佩林的航班在预测市场获利 2.5 万美元

    类似技术在 2008 年曾被用来'预测'莎拉·佩林成为竞选搭档。当时 John McCain 准备宣布副手人选,交易员 Fitzpatrick 等人监控所有竞争者的动向,打电话给其新闻秘书并查航班追踪网站。他们注意到一架从阿拉斯加安克雷奇飞往 McCain 集会地点的湾流公务机,于是买入了冷门的佩林。意外人选公布后,Fitzpatrick 赚了 2.5 万美元。(来源是 BuzzFeed,但这篇文章非常值得一读。)

    paulgb查看英文原文

    A similar technique was used to "predict" Sarah Palin's choice as running mate in 2008. > In 2008, as John McCain prepared to announce his running mate, Fitzpatrick and his fellow trader Joe Schilling were monitoring the movements of all the contenders, calling their press secretaries and checking a flight-tracking website. They noticed a Gulfstream jet from Anchorage, Alaska, bound for the site of a McCain rally, and bought Sarah Palin, a longshot. When the surprise pick was unveiled, Fitzpatrick took home $25,000. Source: http://www.buzzfeed.com/andrewrice/the-fall-of-intrade-and-the-business-of-betting-on-real-life (despite being from buzzfeed, the article is well worth reading)

    相关资源

    从这篇论文所有 HN 讨论帖的全部评论里整理出来的链接。

    代码
    视频与演讲

    历次讨论

    同领域还值得读

    1. 一切的回报率,1870–2015

      The Rate of Return on Everything, 1870–2015

      利用1870–2015年多国的长期数据,估算股票、住房等各类资本的回报率,并将其与经济增长率进行比较。

      经济与社会科学2019 · 487 分 · 260 评进阶 · 约 100 分钟编辑精选
    2. 一切资产的回报率,1870–2015

      The Rate of Return on Everything, 1870–2015

      构建了 1870–2015 年 16 个国家的长周期资产回报数据集,覆盖股票、住房、政府债券和短期票据,系统比较各类资产的真实回报与波动性。

      经济与社会科学2018 · 215 分 · 151 评进阶 · 约 120 分钟编辑精选
    3. 一切资产的回报率,1870–2015(2019)

      The Rate of Return on Everything, 1870–2015 (2019)

      这篇发表于《经济学季刊》的论文构建了 1870–2015 年间 16 个主要发达国家的股票、债券、房地产等资产收益率的长周期数据库。

      经济与社会科学2024 · 190 分 · 148 评进阶 · 约 80 分钟编辑精选
    4. 想法是否越来越难找?

      Are ideas getting harder to find?

      探讨“想法是否越来越难找到”:通过衡量多个领域的“研究生产率”,发现维持相同的增长需要投入指数级增长的研究资源,即研究效率在持续下降。

      经济与社会科学2016 · 182 分 · 106 评入门 · 约 60 分钟编辑精选
    5. 比特币与区块链的经济极限

      The Economic Limits of Bitcoin and the Blockchain

      芝加哥大学 Eric Budish 的论文从经济学角度分析比特币与区块链的根本限制:核心论点是只要攻击网络(双花或砸盘做空)的收益超过挖矿成本,51% 算力攻击就有利可图,而挖矿…

      经济与社会科学2018 · 176 分 · 37 评进阶 · 约 90 分钟编辑精选

    打赏支持

    打赏二维码

    觉得有用,请作者喝杯咖啡。

    WeChat

    关注向阳乔木

    微信公众号「向阳乔木推荐看」

    微信公众号「向阳乔木推荐看」二维码

    微信扫一扫;在微信里打开本页时,长按二维码识别。