Abstract--Learning preferences implicit in the choices humans make is a well studied problem in both economics and computer science. However, most work makes the assumption that humans are acting (noisily) optimally with respect to their preferences. Such approaches can fail when people are themselves learning about what they want. In this work, we introduce the assistive multi-armed bandit, where a robot assists a human playing a bandit task to maximize cumulative reward. In this problem, the human does not know the reward function but can learn it through the rewards received from arm pulls; the robot only observes which arms the human pulls but not the reward associated with each pull. We offer sufficient and necessary conditions for successfully assisting the human in this framework. Surprisingly, better human performance in isolation does not necessarily lead to better performance when assisted by the robot: a human policy can do better by effectively communicating its observed rewards to the robot. We conduct proof-of-concept experiments that support these results. We see this work as contributing towards a theory behind algorithms for humanrobot interaction. I. INTRODUCTION Preference learning [1] seeks to learn a predictive model of human preferences from their observed behavior.

Martin Ford is a futurist and the author of two books: The New York Times Bestselling Rise of the Robots: Technology and the Threat of a Jobless Future (winner of the 2015 Financial Times/McKinsey Business Book of the Year Award and translated into more than 20 languages) and The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future, as well as the founder of a Silicon Valley-based software development firm. His TED Talk on the impact of AI and robotics on the economy and society, given on the main stage at the 2017 TED Conference, has been viewed more than 2 million times. Martin is also the consulting artificial intelligence expert for the new "Rise of the Robots Index" from Societe Generale, underlying the Lyxor Robotics & AI ETF, which is focused specifically on investing in companies that will be significant participants in the AI and robotics revolution. He holds a computer engineering degree from the University of Michigan, Ann Arbor and a graduate business degree from the University of California, Los Angeles. He has written about future technology and its implications for publications including The New York Times, Fortune, Forbes, The Atlantic, The Washington Post, Harvard Business Review, The Guardian, and The Financial Times.

Vivek Goyal isn't a professional photographer, but he and his colleagues have developed an intriguing party trick: they can capture the image of an object completely out of sight. They demonstrated the trick in a windowless room on the Boston University campus, where Goyal works as an electrical engineering professor. In the room, a flat-screen monitor displayed a series of crude drawings created by Goyal's graduate student, Charles Saunders. Among them were several masterpieces: A mushroom that resembles Toad from Mario Kart, a Simpsons-yellow dude wearing a sideways red baseball cap, the red letters "BU" for school pride. These are the images that Goyal and his team wanted to capture while pointing the camera lens in a completely different direction.

In the movie "Terminator 2: Judgment Day," the titular robot says, "My CPU is a neural net processor, a learning computer." The neural network of which that famous robot speaks is actually a real kind of machine learning method. A neural network is a way to depict a complex mathematical formula, organized into layers. This formula can be trained to do things like recognize images for self-driving cars. For example, watch several seconds of a neural network performing object recognition.

Back in the 1980s, Artificial Intelligence (AI) just had to look good in the movies and be able to power fancy talking cars, zappy spaceships and various forms of fantastical cyborgs who would one day roam the planet and possibly destroy the human race. Fast-forward to 2019 and we find ourselves deep in the AI renaissance (or perhaps first'real' birth of AI, rather than any form of rebirth) as we now have the processing power, memory capacity, cloud network breadth and sophisticated algorithmic intelligence to actually apply AI to our lives. But there's a problem -- we (the humans) who build the AI brains need to be able to construct them with a pure enough form of digital DNA such that they stay clean of any form of discriminatory bias. Major cloud networks have already been criticized for employing software that discriminates against women; a well-known search engine has been accused of featuring ethnic bias in results when looking for'unprofessional hairstyles'; an equally well-known social network has been criticized for showing certain job ads only to men; and the list goes on. The question the tech industry must now face is: how to we rid AI of bias in all its forms and ensure fair play for all in the age of computer-driven decision making? One set of answers comes from Women Leading in AI (WLinAI), a network of leaders working in tech, science, politics, business and think tanks – the group is demanding that the UK government takes back control of technology.

Collaborative filtering often suffers from sparsity and cold start problems in real recommendation scenarios, therefore, researchers and engineers usually use side information to address the issues and improve the performance of recommender systems. In this paper, we consider knowledge graphs as the source of side information. We propose MKR, a Multi-task feature learning approach for Knowledge graph enhanced Recommendation. MKR is a deep end-to-end framework that utilizes knowledge graph embedding task to assist recommendation task. The two tasks are associated by cross&compress units, which automatically share latent features and learn high-order interactions between items in recommender systems and entities in the knowledge graph. We prove that cross&compress units have sufficient capability of polynomial approximation, and show that MKR is a generalized framework over several representative methods of recommender systems and multi-task learning. Through extensive experiments on real-world datasets, we demonstrate that MKR achieves substantial gains in movie, book, music, and news recommendation, over state-of-the-art baselines. MKR is also shown to be able to maintain a decent performance even if user-item interactions are sparse.

Music genre recognition based on visual representation has been successfully explored over the last years. Recently, there has been increasing interest in attempting convolutional neural networks (CNNs) to achieve the task. However, most of existing methods employ the mature CNN structures proposed in image recognition without any modification, which results in the learning features that are not adequate for music genre classification. Faced with the challenge of this issue, we fully exploit the low-level information from spectrograms of audios and develop a novel CNN architecture in this paper. The proposed CNN architecture takes the long contextual information into considerations, which transfers more suitable information for the decision-making layer. Various experiments on several benchmark datasets, including GTZAN, Ballroom, and Extended Ballroom, have verified the excellent performances of the proposed neural network. Codes and model will be available at "ttps://github.com/CaifengLiu/music-genre-classification".

Do neutrinos, the elementary particles, have something in common with fake news on social media? The peculiar and positive answer comes from a group of researchers at USI Institute of Computational Science, and it shows how both their behaviour can be represented using the same data structure. Such structure is based on a non-Euclidean geometry and can be studied through a new class of algorithms: the Graph Convolutional Neural Networks (GCNN). Such algorithms are highly complex mathematical models, and the research work carried out by Federico Monti, member of Prof. Michael Bronstein group, earned him the award for the best scientific contribution assigned by ICMLA, the most important international conference in the field. Monti, in collaboration with other colleagues from New York University, Berkeley and Imperial College, had the opportunity to collaborate with the Lawrence Berkley National Laboratory on data acquired by the IceCube Neutrino Observatory at the South Pole.

Artificial Intelligence provides critical information to decision-makers when lives are at stake. But how do we make sure that it doesn't become smarter …