Why we urgently need to measure AI's societal impacts How will artificial intelligence systems change the way we live? This is a tough question: on one hand, AI tools are producing compelling advances in complex tasks, with dramatic improvements in energy consumption, audio processing, and leukemia detection. There is extraordinary potential to do much more in the future. On the other hand, AI systems are already making problematic judgements that are producing significant social, cultural, and economic impacts in people's everyday lives. AI and decision-support systems are embedded in a wide array of social institutions, from influencing who is released from jail to shaping the news we see.

We can see a few interesting things in this table. In a similar talk by Owen at ODSC Boston 2015 titled "Open Source Tools and Data Science Competitions", he again summarized common parameters he uses: We can see some minor differences that may be relevant. Finally, Abhishek Thakur, in his post titled "Approaching (Almost) Any Machine Learning Problem" provided a similar table listing out key XGBoost parameters and suggestions for tuning. The spreads do cover the general defaults suggested above and more. It is interesting to note that Abhishek does provides some suggestions for tuning the alpha and beta model penalization terms as well as row sampling. You can develop and evaluate XGBoost models in just a few lines of Python code.

Capgemini plans to make it easier for companies to embed AI, bots and blockchains into their businesses by launching the Applied Innovation Exchange. The Exchange will give businesses access to Capgemini's network of partners, experts and other resources to help them realise the potential of cutting-edge innovations, the firm said. The global network will offer companies support when they wish to build proof of concepts for the Internet of Things (IoT) and prototypes in specialist sectors such as financial services. "We're thrilled to be opening our latest Applied Innovation Exchange in London, giving business leaders the chance to not only see the latest innovations from a host of exciting companies across the globe, but also explore, here and now, what they could achieve within their own organisations," Rory Burghes, VP of the Exchange at Capgemini UK, said. "To support our group ambition and strategy around cloud and digital, and to go beyond our great delivery capabilities, we are focused on our clients who look to us as a major partner to address the demands on their ability to grow, compete and launch new business models leveraging innovation," Lanny Cohen, global CTO of the platform, added.

Spending thousands of hours playing Grand Theft Auto might have questionable benefits for humans, but it could help make computers significantly more intelligent. Several research groups are now using the hugely popular game, which features fast cars and various nefarious activities, to train algorithms that might enable a self-driving car to navigate a real road. But the stunningly realistic scenery found in Grand Theft Auto and other virtual worlds could help a machine perceive elements of the real world correctly. A technique known as machine learning is enabling computers to do impressive new things, like identifying faces and recognizing speech as well as a person can. But the approach requires huge quantities of curated data, and it can be challenging and time-consuming to gather enough.

Researchers at the University of Leuven in Belgium said a study they conducted showed artificial intelligence could help interpret, and thereby improve, lung function tests used to diagnose long-term lung disease. Results of the study were presented Monday at the European Respiratory Society's International Congress. As part of the study, the researchers used data from 968 people who were undergoing complete lung function testing for the first time. Using a concept called "machine learning", they developed an algorithm that takes into account routine lung function parameters and clinical variables of smoking history, body mass index, and age to make a suggestion for the most likely diagnosis. "We have demonstrated that artificial intelligence can provide us with a more accurate diagnosis in this new study," Wim Janssens, senior author of the study, said.

WASHINGTON – CIA Director John Brennan warned on Sunday that Russia has "exceptionally capable and sophisticated" computer capabilities and that the U.S. must be on guard. When asked in a television interview whether Russia is trying to manipulate the American presidential election, Brennan didn't say. But he noted that the FBI is investigating the hacking of Democratic National Committee emails, and he cited Moscow's aggressive intelligence collection and its focus on high-tech snooping. "I think that we have to be very, very wary of what the Russians might be trying to do in terms of collecting information in a cyber realm, as well as what they might want to do with it," he told CBS' "Face the Nation" on the 15th anniversary of the Sept. 11 attacks. On the terrorism threat, Brennan said the U.S. government is much better now at sharing information.

We have seen how the internet of things and the growing phenomenon of'big data' will throw up major problems for consumers and citizens, problems that have as yet barely been grasped by most policy-makers. In this world of growing complexity, the potential for an unintended consequence becomes greater and greater from machines performing an action that was not anticipated. There are key issues, too, about our reliance on data at a time of massive data generation, data storing and data preservation which have the potential to both obscure results and generate injustices. Perhaps the greatest issue that we now face is caused by our blind faith in machines. We have invested them with certainty and – as we have pointed out – we trust them. Part of the reason for this is an odd confusion that has conflated the machines of the industrial age with the machines of the information age.

There is a lot of hype today about deep learning, a class of multilayer perceptrons with some 5-20 layers featuring convolutional and polling layers. Many blogs [1,2,3] discuss the structure of these networks, there is plenty code published so I won't get into much detail here. Several tech companies had invested a lot of money into this research and everyone has very high expectations on performance of these models. Indeed they've been winning image classification competitions for several years now and media are reporting superhuman performance on some visual classification tasks once in a while. Now just looking at the numbers from ImageNet competition is not really telling us much on how good these models really are, we can only maybe confirm that they are much better than whatever came before them (for that benchmark at least).

It's being pitched as a bionic Olympics. Next month, contestants from across the world using robotic exoskeletons, electronic arms, powered wheelchairs and other tech will take part in the world's first Cybathlon. The event near Zurich is intended to highlight how people with physical disabilities can be helped by novel technological aids in their everyday activities. Fifty teams from across the world will take part, including six from the UK. Events include slicing bread, going upstairs and putting out the washing.

The process of dynamic state estimation (filtering) based on point process observations is in general intractable. Numerical sampling techniques are often practically useful, but lead to limited conceptual insight about optimal encoding/decoding strategies, which are of significant relevance to Computational Neuroscience. We develop an analytically tractable Bayesian approximation to optimal filtering based on point process observations, which allows us to introduce distributional assumptions about sensor properties, that greatly facilitate the analysis of optimal encoding in situations deviating from common assumptions of uniform coding. Numerical comparison with particle filtering demonstrate the quality of the approximation. The analytic framework leads to insights which are difficult to obtain from numerical algorithms, and is consistent with biological observations about the distribution of sensory cells' tuning curve centers.