Data Science newsletter – January 7, 2020

GROUP CURATION: N/A

In spite of the fact that it’s possible the development rate for Chinese shopper spending will be somewhat lower in 2019 than in 2018, the Chinese keep on expanding their spending by a significant edge and are anxious to pay for things with a solid offer. McKinsey’s most recent China Consumer Report released this month highlights five trends for 2020:

1. Young, free-spending consumers in lower-tier cities are today’s growth engine.

Summarizing text is a task at which machine learning algorithms are improving, as evidenced by a recent paper published by Microsoft. That’s good news — automatic summarization systems promise to cut down on the amount of message-reading enterprise workers do, which one survey estimates amounts to 2.6 hours each day.

Not to be outdone, a Google Brain and Imperial College London team built a system — Pre-training with Extracted Gap-sentences for Abstractive SUmmarization Sequence-to-sequence, or Pegasus — that leverages Google’s Transformers architecture combined with pretraining objectives tailored for abstractive text generation. They say it achieves state-of-the-art results in 12 summarization tasks spanning news, science, stories, instructions, emails, patents, and legislative bills, and that it shows “surprising” performance on low-resource summarization, surpassing previous top results on six data sets with only 1,000 examples.

Finding a parking place in busy urban areas is a major headache for freight delivery drivers.

But help for them may be available through artificial intelligence and machine learning in an app being developed at Pacific Northwest National Laboratory in Richland.

The Department of Energy lab is using $1.5 million from DOE’s Office of Energy Efficiency and Renewable Energy’s Vehicle Technologies Office to develop, test and improve technologies aimed at cutting driver time.

Data is fast becoming our most valuable commodity. But, beyond just where we store data, our ability to realize the benefits of this commodity is restricted by the availability of people able to work with it. With 2.5 quintillion bytes of potentially unstructured and varying quality data coming from continuously evolving sources on a daily basis, it’s important we think beyond the number of hands we have to do this work, and instead think about how we do this work.

Current solutions to data management will need to evolve to meet new requirements. Automating tasks with the help of machine learning and AI can help with data analysis. AI processing of data will make the 5G data influx much more manageable for organizations.

The model for consuming data could go the way of the music industry. We’ve moved from vinyl to tapes to CDs and now to subscription-based streaming services, which employ machine learning to recommend songs we might be interested in based on what we’ve been listening to. We’ve already started to see data consumption models move in this direction.

1. Financial knowledge is optional — A big difference between Renaissance and other quant funds is that their team consists of scientists, not Wall Street folks. With zero finance background, they treat financial data like the scientific/text data they used to experiment on. An embarrassingly funny scene happened when Bob Mercer was asked how they made so much money with their models. He replied that “sometimes it tells us to buy Chrysler, sometimes it tells us to sell.” This was when Chrysler was no longer trading after having been acquired — it shows how little Renaissance needs to know about the underlying fundamentals of the companies, even their names!

“We solved mode collapse and log-likelihood evaluation for GANs (https://arxiv.org/abs/1910.04302 ) … I still see papers on GANs coming out that don’t build on this advance.” And many more.

We enter a new decade with great expectations of prosperity, as machine learning finds its place in traditional industries from manufacturing to the arts. Yet we face important questions about how to use it without causing harm through careless data collection, slipshod system design, or the limits of our ability to see around the next corner. In this special issue of The Batch, some of the brightest lights in AI express their hopes for 2020.

Dominic Cummings, a senior adviser to UK prime minister Boris Johnson, has said he wants the UK government to hire “weirdos and misfits with odd skills” to apply science to the civil service. While primarily a quirky job ad, his blog post also offers a glimpse into how he sees scientific research transforming the government.

As well as listing categories of people he would like to hire – including mathematicians and physicists – the blog post also focuses on the utility of data science, artificial intelligence and the “science of prediction”.

But does his vision make sense? Can policy-making really be improved by building digital models of reality, or applying machine-learning to government data, as Cummings appears keen on?

Al Gore’s 1992 forecast of a Digital Earth — where satellites beam data to reveal all the planet’s environmental dynamics – has gained momentum with the publication of the Manual of Digital Earth last month. The major anthology is sponsored by the Chinese Academy of Sciences. It’s a mark of the importance China attaches to what is now a United Nations-led project named the Global Earth Observation System of Systems (GEOSS).

GEOSS seems like medical science’s worldwide collaborations to map the human genome and the human brain – but at a much bigger magnitude. Scientists want to data-visualise the whole Earth. The project’s scope ranges from deep subterranean core samples, volcanic tremors, ocean surface temperatures, flooding and solar storms to urban populations, migrations and sprawls.

A University of Waterloo study had 10 men and 10 women interact with Amazon Alexa, Apple Siri, and Google Assistant and then asked the participants questions about the voice assistant personalities. Amazon, Apple, and Google have each gone to great lengths to create personas for their assistants to make them more relatable for consumers. Edward Lank, a professor in Waterloo’s School of Computer Science, suggests these efforts are paying off in at least one way: “People are anthropomorphizing these conversation agents which could result in them revealing information to the companies behind these agents that they otherwise wouldn’t.”

Your local electricity provider uses AI and so does the person who takes those goofy real-estate agent pictures you see on park benches. Everything is AI — an axiom that’ll become even truer in 2020.

We solicited predictions for the AI industry over the next year from a panel of experts, here’s what they had to say.

Neuroscientific research and tools now allow us to examine more deeply some of the ways in which individuals experience spaces and natural elements. This knowledge can greatly add to, and shift, the priorities and direction of urban design and planning.

What do we mean by ‘nature’?

A large body of research has compellingly shown that “nature” in its many forms and contexts can have direct benefits on mental health. Unfortunately, the extent and diversity of natural habitats in our cities are decreasing rapidly.

Adobe CTO Abhay Parasnis sees a shift happening.

A shift in how people share content and who wants to use creative tools. A shift in how users expect these tools to work — especially how much time they take to learn and how quickly they get things done.

I spoke with Parasnis in December to learn more about where Adobe’s products are going and how they’ll get there — even if it means rethinking how it all works today.

The Trump administration took measures on Friday to crimp exports of artificial intelligence software as part of a bid to keep sensitive technologies out of the hands of rival powers like China.

Under a new rule which goes into effect on Monday, companies that export certain types of geospatial imagery software from the United States must apply for a license to send it overseas except when it is being shipped to Canada.

“They’re high enough where you couldn’t shoot one anyway, but they’re low enough that they’re a nuisance,” said Dawn George, who lives near Wray, Colo., and whose Border collie has barked at the drones when they fly over her property.

Ms. George said she had heard wild speculation about who might be responsible for the flights — the government? a cartel? a gas company? — and feared they would never know the truth.

“All the sudden, it’s just going to stop and we’re not going to have answers,” Ms. George said. “And that’s very unsettling to a lot of people. It’s the fear of the unknown.”