Greg Linden may not be a household name, but he changed the way we interact with culture and transformed retail forever. An engineer at Amazon in the late 1990s, Linden worked on a curious problem: how to recommend books without human intervention. Until then Amazon relied on editors who wrote hundreds of reviews every year. It was a costly and time-consuming process. Automating recommendations proved trickier than anyone expected.

Volvo has decided to join the ranks of automakers with offices in Silicon Valley. The Swedish car company is in the process of opening a research center in Mountain View, Lex Kerssemakers, CEO of Volvo's U.S. division, said in an interview. The company is hiring some 70 engineers for the office, he said. Volvo, which is owned by Chinese automaker Geely but operates largely independently, has had an office in Camarillo for about 30 years that focused on car design, Kerssemakers said. Within the past three to four years, engineers based in that office also started to work on car infotainment systems, he said.

This paper proposes a general adaptive procedure for budget-limited predictor design in high dimensions called two-stage Sampling, Prediction and Adaptive Regression via Correlation Screening (SPARCS). SPARCS can be applied to high dimensional prediction problems in experimental science, medicine, finance, and engineering, as illustrated by the following. Suppose one wishes to run a sequence of experiments to learn a sparse multivariate predictor of a dependent variable $Y$ (disease prognosis for instance) based on a $p$ dimensional set of independent variables $\mathbf X=[X_1,\ldots, X_p]^T$ (assayed biomarkers). Assume that the cost of acquiring the full set of variables $\mathbf X$ increases linearly in its dimension. SPARCS breaks the data collection into two stages in order to achieve an optimal tradeoff between sampling cost and predictor performance. In the first stage we collect a few ($n$) expensive samples $\{y_i,\mathbf x_i\}_{i=1}^n$, at the full dimension $p\gg n$ of $\mathbf X$, winnowing the number of variables down to a smaller dimension $l < p$ using a type of cross-correlation or regression coefficient screening. In the second stage we collect a larger number $(t-n)$ of cheaper samples of the $l$ variables that passed the screening of the first stage. At the second stage, a low dimensional predictor is constructed by solving the standard regression problem using all $t$ samples of the selected variables. SPARCS is an adaptive online algorithm that implements false positive control on the selected variables, is well suited to small sample sizes, and is scalable to high dimensions. We establish asymptotic bounds for the Familywise Error Rate (FWER), specify high dimensional convergence rates for support recovery, and establish optimal sample allocation rules to the first and second stages.

This might be one of the most dangerous yet breath-taking car journeys in the world. A new drone video has emerged showing the dangerous mountain tunnel chiselled out of a vertical cliff in Taihang Mountains. The footage, posted to Youtube by CCTV News yesterday, is believed to capture the awe-inspiring Guoliang tunnel in central China's Henan Province. The formidable Guoliang tunnel was hand-carved by 13 men from an ancient village perched on top of the cliff at an altitude of over 1,700 metres (5,500 feet). The 1,250-metre-long (4,100 feet) passage is the only way to drive to the 600-year-old Guoliang village, which perches on top of the 200-metre-tall (656 feet) cliff.

Welcome back to Across the Network -- Lab41's weekly look at what is going on in the world of AI. As always these are all links that I pulled from the Lab41 Slack channels. A Neural Network for Machine Translation -- I have a 3-year old son, who, thanks to his mother (whose family is from Taiwan) and our Chinese au-pair, speaks Mandarin fluently. And while I'm proud of this fact (and a bit miffed that he is so much more capable at picking up language than me), his fluency has required me to become a regular user of Google Translate. So I was excited to see the details behind the latest technology being used by the Google Translate team.

At the two-day RoboBusiness Conference, about 2,000 people were serenaded with lullabies and Disney tunes, including "Let It Go" from the hit film "Frozen," by a human-like robot designed to comfort senior citizens and autistic children. And next to a man-size robot that can drive a motorcycle 190 mph around a race track, a half-dozen ant-size robots quickly scurried about a miniature factory floor. "In five years, could you imagine what this conference is going to look like?" "There are going to be 8-foot robots walking all around us, talking to us, some of them maybe being smarter than us." The 12th annual conference, which wrapped up Thursday, illustrated how the focus of robotics is shifting from industrial uses to consumer products. That's especially true at a time when drones, self-driving cars and police robots that carry bombs are making news.

Nvidia Corp. has unveiled two more graphics processing unit (GPU) chips aimed at the fast-growing branch of artificial intelligence called deep learning. At the GPU Technology Conference currently underway in Beijing, Nvidia CEO Jen-Hsun Huang (pictured) introduced the Tesla P4 and Tesla P40 GPUs that form part of the Pascal architecture-based deep learning platform. The company had officially entered the deep learning market in April with the announcement of the 15-billion transistor Tesla P100 chip, aimed at deep learning. "With the Tesla P100 and now Tesla P4 and P40, Nvidia offers the only end-to-end deep learning platform for the data center, unlocking the enormous power of AI for a broad range of industries," said Ian Buck, general manager of accelerated computing at Nvidia. While the Tesla P100 focuses on training tasks, the Tesla P4 and P40 has been designed specifically for inferencing.

Shares of tech companies rebounded after a stellar stock-market debut for a software company quieted fears about lack of appetite for new tech issues. Nutanix shares more than doubled on its stock-market debut after the maker of software for data management priced its initial public offering above expectations. Japanese conglomerate Softbank's Chief Executive Masayoshi Son said the company plans to invest around 4.5 billion in South Korea's technology sector over the next decade, expressing interest in emerging fields of smart robots, the "Internet of Things" and artificial intelligence.

Apparently, director Tim Burton would be fine if you retitled his movie "Miss Peregrine's Home for Peculiar Children but Only if They're White." In an interview with the website Bustle, Burton was asked why, given the pervasive, ongoing discussion of diversity in Hollywood, the overwhelming majority of the characters (and, hence, actors) in his latest fantasy film are white. "Nowadays, people are talking about it more," Burton said. He went on to talk about how, as a kid, he would've been dismayed to see an Asian kid or a black kid on "The Brady Bunch," or more white actors in blaxploitation movies. Burton's statements are just the latest from a celebrated, veteran filmmaker unable to wrap his mind around why diversity matters.

Rebuilding fine motor skills after a stroke takes intensive therapy involving repeated attempts to use the affected hand, several times a day, day after day. Some involve moving the affected fingers with the other hand until new brain pathways for hand control develop. Zain Samdani, a 16-year-old from Saudi Arabia and a finalist in the 2016 Google Science Fair demonstrated a different approach at the finalist showcase on Tuesday. Samdani, who says he'd seen family members struggling with hand rehab, built an exoskeleton out of 3D-printed segments. He connected that to a glove, wired to control the robotic device.