Goto

Collaborating Authors

 Europe


Distributed Optimization of Multi-Class SVMs

arXiv.org Machine Learning

Training of one-vs.-rest SVMs can be parallelized over the number of classes in a straight forward way. Given enough computational resources, one-vs.-rest SVMs can thus be trained on data involving a large number of classes. The same cannot be stated, however, for the so-called all-in-one SVMs, which require solving a quadratic program of size quadratically in the number of classes. We develop distributed algorithms for two all-in-one SVM formulations (Lee et al. and Weston and Watkins) that parallelize the computation evenly over the number of classes. This allows us to compare these models to one-vs.-rest SVMs on unprecedented scale. The results indicate superior accuracy on text classification data.


Chatting with robots

#artificialintelligence

In the world of e-commerce, providing an outstanding level of customer support can be quite a challenge, especially in terms of scalability. Customers usually go online at night, when there's no customer support available. Furthermore, when they all get in touch with the service desk at the same time they are clogging it up, and often the support they need is related to a frequently asked question. Automating this process allows the organization to provide a far better service to its customers, regardless of the number of customers calling, chatting or emailing at the same time and regardless of the moment they decide to do so. Craftworkz, a Belgian start-up company in the Cronos Group, located in Leuven, builds chatbots to automate customer support.


AI, virtual assistants and chat bots before, now and in the future - Fluxx

#artificialintelligence

I'm taking the next step in the smart world here at Fluxx, and focusing in on AI, smart chat bots and stupid virtual assistants to hopefully find some interesting examples on what's out there. One of the hottest topics of the last few years has been AI in all its forms, everything from simple female-named chatting tools, to domesday predictions that AI will kill us all, or at least make us fat and obsolete in the workplace. So far what we've seen and been able to play with has mostly been in the form of chat bots, helping us navigate over-complicated websites or get some very limited customer service help. November 6, 2001 "Treehouse of Horror XII" was aired, with Pierce Brosnan starring as Ultrahouse 3000. A smart building AI that becomes attracted to Marge and decides to get rid of Homer, attempting to kill him by dumping him into the dining room table's garbage disposal -- many of the common fears of what an AI eventually will do to us. Some of you might remember that a few years back it felt like most service providers started using Automated online assistants like Anna from IKEA here, often used by banks or insurance companies to provide visitors with another form of customer service.


How Analytics and AI Are Driving the Subscription E-Commerce Phenomenon

#artificialintelligence

Box subscription companies are reinventing retail using personalization and artificial intelligence. In recent years, amid a lackluster shopping environment, an unheralded retail phenomenon has taken off quite dramatically -- with growth rates exceeding 1,000%. The startups representing this evolving segment are collectively grouped under the label of subscription e-commerce. With such evocative names as BarkBox, Birchbox, Blue Apron, Harry's, OwlCrate, Trunk Club, and Winc, these online companies mail monthly boxes containing specially curated items in beauty, fashion, food, personal grooming, and pet products priced between $10 and $80 per box right to their subscribers' doorsteps. Attention was brought to this category when consumer giant Unilever snapped up one of the best-known startups -- Dollar Shave Club -- in July 2016 for an eye-popping $1 billion -- five times its annual revenue.


Chipmakers Are Racing To Build Hardware For Artificial Intelligence

#artificialintelligence

In recent years, advanced machine learning techniques have enabled computers to recognize objects in images, understand commands from spoken sentences, and translate written language. But while consumer products like Apple's Siri and Google Translate might operate in real time, actually building the complex mathematical models these tools rely on can take traditional computers large amounts of time, energy, and processing power. As a result, chipmakers like Intel, graphics powerhouse Nvidia, mobile computing kingpin Qualcomm, and a number of startups are racing to develop specialized hardware to make modern deep learning significantly cheaper and faster. The importance of such chips for developing and training new AI algorithms quickly cannot be understated, according to some AI researchers. "Instead of months, it could be days," Nvidia CEO Jen-Hsun Huang said in a November earnings call, discussing the time required to train a computer to do a new task.


AI Podcast: Where Is Deep Learning Going Next? NVIDIA Blog

#artificialintelligence

We'll know AI really works when we hardly notice it at all, according to Bryan Catanzaro, a key figure in the field. "AI gets better and better until it kind of disappears into the background," says Catanzaro -- NVIDIA's head of applied deep learning research -- in conversation with host Michael Copeland on this week's edition of the new AI Podcast. "Once you stop noticing that it's there because it works so well -- that's when it's really landed." Bryan's been in AI since the beginning. Or, as Michael says, as "about as long as it has really worked."


Five Fast-Growing British Businesses To Watch In 2017

Forbes - Tech

Even the biggest businesses start out small. The sale last month of Skyscanner, the travel industry start-up, is just another example of British entrepreneurial success, built over a number of years – and defies those who bemoan the country's failure to produce big winners. But which are the businesses to watch in 2017? Well, while picking winners from smaller companies is fraught with difficulties, here are five young British companies tipped for big things over the year ahead. Captify is an insights-driven advertising technology company founded in 2011 that, via its purpose-built Search Intelligence platform, analyses more than 15 billion online searches each month.


Are World Of Warcraft, Battle.net Down? Blizzard Says DDoS Attack Affecting Its Games

International Business Times

UPDATE: 12:52 p.m. EST - Following user reported outages across the globe, Blizzard Entertainment said Wednesday a Distributed Denial of Service, or DDoS, attack has stopped and encouraged users of its games to attempt to log back into their systems. The DDOS attacks against network providers that we were monitoring have ended. If you are still unable to log in try https://t.co/Mce1fu5ZI9 Original story - Blizzard Entertainment, the maker of wildly popular video games like "World of Warcraft" and "Diablo," announced Wednesday that a Distributed Denial of Service, or DDoS, attack had affected connections to its games. We are currently monitoring a DDOS attack against network providers which is affecting latency/connections to our games.


Half surface water in some countries has been lost since 1980s

New Scientist

More of Earth's surface is covered by liquid water now than three decades ago. But some countries in Central Asia and the Middle East have lost more than half their surface water, satellite images show. There have also been losses in the US and Australia. The changes are mainly a result of activities such as irrigation and dam building, but climate change is playing a part too. Previous studies of surface water have largely relied on how much water countries estimate they have..


The AI that could uncover the secret of eternal youth

#artificialintelligence

Scientists have revealed a new plan to find the key to eternal youth – and artificial intelligence will be leading the way. Using computer simulations to screen hundreds of compounds, researchers have developed a tool that can identify geroprotectors, the substances responsible for extending healthy life. GeroScope can compare changes in the cells of young and old patients and search for the drugs that counteract the processes. The project is led by scientists from Moscow Institute of Physics and Technology and Insilico Medicine Inc, commissioned by the Center for Biogerontology and Regenerative Medicine. According to the researchers, using computer modelling techniques can help to cut down time and cost in the development of age-combating drugs.