Nightfall’s Up to Speed on AI and Deep Learning 1/21/20

New developments

• Why Healthcare Machine Learning Adoption is Skyrocketing
  (DevPro Journal)
  Chandra Kalle, Senior Engineering Director of health AI startup LeanTaaS, speaks with DevPro Journal about the utility and opportunities for machine learning in healthcare settings.
• Amazon’s AutoGluon produces AI models with as little as 3 lines of code
  (VentureBeat)
  Amazon developed AutoGluon, an open source library designed to enable developers to write AI-imbued apps with only a few lines of code. It publicly launched this month.
• Coral is Google’s quiet initiative to enable AI without the cloud
  (The Verge)
  Google Coral, a stealth initiative, promises to make computers faster and more secure with on-device AI. Although the project actually has its roots in Google’s “AIY” range of do-it-yourself machine learning kits, the long-term focus is on enterprise customers in industries like the automotive world and health care.
• Ancestry turned to AI to bring down cloud costs
  (CIO Dive)
  Ancestry spent two years migrating a database of over 20 million members away from data centers and into Amazon Web Services. After the move, the company turned attention to optimizing its presence in the cloud with Opsani, an AIOps company that relies on machine learning to manage cloud workloads.

Research

• Smart Data based Ensemble for Imbalanced Big Data Classification
  (arXiv)
  Big Data scenarios pose a new challenge to traditional data mining algorithms since they are not prepared to work with such amount of data. Smart Data refers to data of enough quality to improve the outcome from a data mining algorithm. Existing data mining algorithms inability to handle Big Datasets prevents the transition from Big to Smart Data. Experiments carried out in 21 Big Datasets have proved that the authors’ ensemble classifier outperforms classic machine learning models with an added data balancing method, such as Random Forests.

• Identifying Table Structure in Documents using Conditional Generative Adversarial Networks
  (arXiv)
  Hierarchically-related data is rendered as tables, and extracting information from tables in such documents presents a significant challenge. The authors propose a top-down approach, first using a conditional generative adversarial network to map a table image into a standardized skeleton table form denoting approximate row and column borders without table content, then deriving latent table structure using xy-cut projection and Genetic Algorithm optimization. 
• Modeling and solving the multimodal car- and ride-sharing problem
  (arXiv)
  The authors introduce the multimodal car-and ride-sharing problem (MMCRP), in which a pool of cars is used to cover a set of ride requests, while uncovered requests are assigned to other modes of transport (MOT). The problem can be formulated as a vehicle scheduling problem. In order to solve the problem, an auxiliary graph is constructed in which each trip starting and ending in a depot, and covering possible ride-shares, is modeled as an edge in a time-space graph. They propose a two-layer decomposition algorithm based on column generation, where the master problem ensures that each request can only be covered at most once, and the pricing problem generates new promising routes by solving a kind of shortest path problem in a time-space network.
• Beyond Near- and Long-Term: Towards a Clearer Account of Research Priorities in AI Ethics and Society
  (arXiv)
  One way of carving up the broad "AI ethics and society" research space that has emerged in recent years is to distinguish between "near-term" and "long-term" research. While such ways of breaking down the research space can be useful, we put forward several concerns about the near/long-term distinction gaining too much prominence in how research questions and priorities are framed. We highlight some ambiguities and inconsistencies in how the distinction is used and argue that while there are differing priorities within this broad research community, these differences are not well-captured by the near/long-term distinction. 
• Should Artificial Intelligence Governance be Centralised? Design Lessons from History
  (arXiv)
  The authors draw on the history of other international regimes to identify advantages and disadvantages in centralizing AI governance. Some considerations, such as efficiency and political power, speak in favor of centralization. Conversely, the risk of creating a slow and brittle institution speaks against it, as does the difficulty in securing participation while creating stringent rules. Other considerations depend on the specific design of a centralized institution. Centralization entails trade-offs and the details matter.
• The Penetration of Internet of Things in Robotics: Towards a Web of Robotic Things
  (arXiv)
  Some of the benefits of IoT in robotics have been discussed in related work. This paper moves one step further, studying the actual current use of IoT in robotics, through various real-world examples encountered through bibliographic research. The paper also examines the potential of WoT, together with robotic systems, investigating which concepts, characteristics, architectures, hardware, software and communication methods of IoT are used in existing robotic systems, which sensors and actions are incorporated in IoT-based robots, as well as in which application areas. Finally, the current application of WoT in robotics is examined and discussed.

AI and ML in Society

• The Problem with Hiring Algorithms
  (Machine Learning Times)
  Brian Gallagher of NYU’s Ethical Systems summarizes the status of facial recognition and other types of analytics used to assess potential employees in interviews as well as whether or not they function as intended.

• We’re fighting fake news AI bots by using more AI. That’s a mistake.
  (MIT Technology Review)
  Facebook and others are battling complex disinformation with AI-driven defenses. But this can only get us so far, argues an expert on high-tech propaganda.
• Rethinking Business Strategy in the Age of AI
  (Harvard Business)
  For the first time in 100 years, new technologies such as artificial intelligence are causing firms to rethink their competitive strategy and organizational structure, say Marco Iansiti and Karim R. Lakhani, authors of the new book Competing in the Age of AI.
• Christiana Care offers tips to 'personalize the black box' of machine learning
  (Healthcare IT News)
  At HIMSS20, its Chief Health Information Officer will show how the health system is simplifying AI models so care managers can better understand them – and be more likely to use them.

Receive the next edition of Up to Speed on AI and Deep Learning directly via email by subscribing here.