2016-08-12 TalkingData 锐眼看世界:DeepMind 试图理解人脑工作机制,BAH 如何培养数据科学家,Garner 发布 IoT Top10 技术

锐眼视点:DeepMind 正在和科学家一家试图解密人脑是如何做计划和决策的;Booz Allen Hamilton 如何在内部从零开始培养数据科学家;Tesla Autopilot 8 小时体验报告;Garner 发布未来两年最重要的 10 个 IoT 技术。

TD 精选

DeepMind 试图理解人们看到地铁路线图时大脑时大脑是如何工作的

原文链接:Google's AI gurus ran tests to try and understand how the human brain works on a subway
Google DeepMind cofounder and CEO Demis Hassabis

DeepMind 和牛津大学、伦敦大学学院的科学家合作,尝试理解人脑在看地铁路线图时是如何工作的。志愿者会被要求在一个虚拟地铁网络中计划一个行程,与此同时,大脑会接收 MRI 扫描。这些扫描数据可以揭示人脑的哪些部分负责制定计划和作出决定。


Booz Allen Hamilton 如何从零开始培养数据科学家

原文链接:Booz Allen Hamilton Builds Data Scientists From Scratch

Booz Allen Hamilton Inc. 自豪于他们可以为自己的客户提供最好的科技和能力。这家提供管理和科技咨询的百年老店拥有超过 22,000 名雇员,管理着具有战略重要性的项目——很多面向政府部门——这些需要高素质的计算机科学家、工程师、分析师以及其他科技专家。Aimee George Leary, strategic talent officer and senior vice president at Booz Allen Hamilton 提到:

“Having talent with the skills to innovate is a driver for our business,”

为了应对自身对高素质数据科学家的迫切需求,Booz Allen Hamilton 推出了 Tech Tank,面向公司内部具备自身技术角色基础能力但需要培训才能更进一步的初级员工的一年期学习发展项目。公司会安排优秀员工和资深员工帮助参与计划的员工提高技能水平,成长为优秀的数据科学家。


业界新闻

体验了 8 小时 Tesla 无人驾驶的车主有话说

原文链接:I Just Drove Eight Hours on Tesla Autopilot and Lived to Tell the Tale
Autopilot

正常情况下,Autopilot 可以从静止开始加速,并保持在自己的车道内,并且能自动刹车避免碰撞,加减速体验很好,然而测试车辆还是差点和另一辆 SUV 追尾。

该车主提到,不要将 Autopilot 看作自动驾驶,而应该视其为巡航控制系统,所以,要随时准备接管车辆。文中提到了很多一手体验,很值得一读。


媒体视角

Gartner's Top 10 Internet Of Things Technologies For 2017 & 2018

原文链接:Gartner's Top 10 Internet Of Things Technologies For 2017 & 2018

Garner 发布了未来两年最重要的 10 个 IoT 技术:

IoT Security – Gartner predicts that hardware and software advances will make IoT security a fast-evolving area through 2021 and the skills shortage today will only accelerate. Enterprises need to begin investing today in developing this expertise in-house and also begin recruitment efforts. As many security problems are the result of poor design, implementation and lack of training, expect to see market leaders adopting IoT investing heavily in these areas.
IoT Analytics – IoT analytics require entirely new algorithms, architectures, data structures and approaches to machine learning if organizations are going to get the full value of the data captured, and knowledge created. Distributed analytics architectures the capitalize on pervasive, secure Internet of Things (IoT) network architectures will eventually become knowledge sharing networks. For more information on how Toyota accomplished this, please see the research completed by Dr. Jeffrey Dyer and Dr. Nobeoka, Creating and Managing A High-Performance Knowledge-Sharing Network: The Toyota Case (Dyer, Nobeoka, 2000).
**IoT Device Management **– The challenges of enabling technologies that are context, location, and state-aware while at the same time consistent with data and knowledge taxonomies is an area Gartner believes will see significant innovation in the next few years. IoT Device Management will most likely break the boundaries of traditional data management and create data structures capable of learning and flexing to unique inbound data requirements over time.
Low-Power, Short-Range IoT Networks – Low-power, short-range networks will dominate wireless IoT connectivity through 2025, far outnumbering connections using wide-area IoT networks.
Low-Power, Wide-Area Networks – According to Gartner, traditional cellular networks don’t deliver a proper combination of technical features and operational cost for those IoT applications that need wide-area coverage combined with relatively low bandwidth, good battery life, low hardware and operating cost, and high connection density.
**IoT Processors **– Gartner predicts that low-end 8-bit microcontrollers will dominate the IoT through 2019 and shipments of 32-bit microcontrollers will overtake the 8-bit devices by 2020. It’s interesting to note that Gartner doesn’t see 16- bit processors ever attaining critical mass in IoT applications.
IoT Operating Systems – Minimal and small footprint operating systems will gain momentum in IoT through 2020 as traditional large-scale operating systems including Windows and iOS are too complex and resource-intensive for the majority of IoT applications. It’s been my experience that these operating systems are excellent at exception- and event-driven tasks and can a few support the essential of multithreading as well.
**Event Stream Processing **- Gartner predicts that some IoT applications will generate extremely high data rates that must be analyzed in real time. Systems creating tens of thousands of events per second are common, and millions of events per second can occur in some telecom and telemetry situations. To address such requirements, distributed stream computing platforms (DSCPs) have emerged.
**IoT Platforms **– According to Gartner IoT platforms bundle infrastructure components of an IoT system into a single product. The services provided by such platforms fall into three core categories. These include low-level device control and operations such as communications, device monitoring and management, security, and firmware updates; IoT data acquisition, transformation and management; and IoT application development, including event-driven logic, application programming, visualization, analytics and adapters to connect to enterprise systems.
IoT Standards and Ecosystems – Although ecosystems and standards aren’t precisely technologies, most eventually materialize as application programming interfaces (APIs). Standards and their associated APIs will be essential because IoT devices will need to interoperate and communicate, and many IoT business models will rely on sharing data between multiple devices and organizations.


你可能感兴趣的:(2016-08-12 TalkingData 锐眼看世界:DeepMind 试图理解人脑工作机制,BAH 如何培养数据科学家,Garner 发布 IoT Top10 技术)