When director Stanley Kubrick released 2001: A Space Odyssey 50 years ago, he intended it to be a "nonverbal experience." The movie's dialogue was sparse, and it relied heavily on visuals and score. It was, Kubrick told Playboy in 1968, a subjective film, meant to reach audiences "at an inner level of consciousness, just as music does." The same cannot be said of its 1976 comic book adaptation. Marvel's oversized Treasury Edition, which was written, drawn, and edited by the late Jack Kirby--legendary co-creator of Captain America, the X-Men, Black Panther, and dozens of others--left little up to interpretation.

Just because Netflix's films got shut out of competition at Cannes doesn't mean the streaming titan stayed home. The company won a worldwide (except for China) distribution deal worth $30 million for the animated film Next Gen with a star-studded cast about a pair of "two unlikely friends in a world filled with robots," according to Deadline. Not much else is known about the film, although Deadline surmised it's one of the biggest deals to come out of the film festival thus far. The voice cast includes Charlyne Yi, Michael Peña, Constance Wu, Jason Sudeikis and David Cross. It's unclear when Netflix will distribute the film, but we know when it's coming out in China: Alibaba and Wanda, which won rights to distribute in the country, will give it a wide release this summer.

With the increase of machine learning usage by industries and scientific communities in a variety of tasks such as text mining, image recognition and self-driving cars, automatic setting of hyper-parameter in learning algorithms is a key factor for achieving satisfactory performance regardless of user expertise in the inner workings of the techniques and methodologies. In particular, for a reinforcement learning algorithm, the efficiency of an agent learning a control policy in an uncertain environment is heavily dependent on the hyper-parameters used to balance exploration with exploitation. In this work, an autonomous learning framework that integrates Bayesian optimization with Gaussian process regression to optimize the hyper-parameters of a reinforcement learning algorithm, is proposed. Also, a bandits-based approach to achieve a balance between computational costs and decreasing uncertainty about the Q-values, is presented. A gridworld example is used to highlight how hyper-parameter configurations of a learning algorithm (SARSA) are iteratively improved based on two performance functions.

ABSTRACT We present a feature engineering pipeline for the construction of musical signal characteristics, to be used for the design of a supervised model for musical genre identification. The key idea is to extend the traditional two-step process of extraction and classification with additive stand-alone phases which are no longer organized in a waterfall scheme. The whole system is realized by traversing backtrack arrows and cycles between various stages. In order to give a compact and effective representation of the features, the standard early temporal integration is combined with other selection and extraction phases: on the one hand, the selection of the most meaningful characteristics based on information gain, and on the other hand, the inclusion of the nonlinear correlation between this subset of features, determined by an autoencoder. The results of the experiments conducted on GTZAN dataset reveal a noticeable contribution of this methodology towards the model's performance in classification task. Index Terms-- Musical signal, genre classification, feature extraction and selection, information gain, autoencoder 1. INTRODUCTION One of the current subjects of research in Computer Science and Engineering concerns the enhancement of machines with abilities which are related to the human perception of the environment.

Do you think Nikita from New York and Melissa from Memphis would buy the same kind of clothes or be attracted to the same top? It's hard to argue that they will. Yet, H&M and hundreds of other retailers choose to bring the same "new" product line and replicate the same store format, in all outlets across the country. However, if the Wall Street Journal's reports are to be believed, H&M has learned to question this practice. Instead, the retailer will use big data and artificial intelligence (AI) to radically transform how it manages stores in each location to better engage with and entice shoppers.

Judea Pearl, chancellor's professor of computer science and statistics at UCLA, has written his first book intended for a general audience, "The Book of Why: The New Science of Cause and Effect." The book, which was written with co-author Dana Mackenzie, explores causality -- the study of cause and effect -- from its origins to its applications at the leading edges of science. Pearl, a UCLA faculty member since 1970, received the 2011 A.M. Turing Award, considered the "Nobel Prize" in computing, for his landmark work in processing information under uncertainty. His new book will be published on May 15. That same day, Pearl will deliver a talk at the Charles E. Young Research Library as part of the UCLA Library Writer Series.

Many e-commerce and retail companies are leveraging the power of data and boosting sales by implementing recommender systems on their websites. In short, these systems aim to predict users' interests and recommend items that quite likely are interesting for them. Data required for recommender systems stems from explicit user ratings after watching a movie or listening to a song, from implicit search engine queries and purchase histories, or from other knowledge about the users/items themselves. Sites like Spotify, YouTube or Netflix use that data in order to suggest playlists, so-called Daily mixes, or to make video recommendations, respectively. In this blog post, we'll describe the broad types of the most popular recommender systems and give insights into how they work, going through a few examples. To give some motivation on the subject and help decide whether it's a worthwhile investment, we'll point to some real-life case studies, talk about the high level requirements for implementing recommender systems, and discuss how they can be evaluated fairly.

Almost everyone knows the story of HAL 9000, the killer supercomputer in Stanley Kubrick's landmark film 2001: A Space Odyssey, whose 50th anniversary will be celebrated on May 12, 2018 at the 71st Cannes Film Festival. In an intriguing scheduling coincidence, IBM, Kubrick's partner during the filming of A Space Odyssey, and Airbus have just unveiled the CIMON (Crew Interactive Mobile Companion) project, an "intelligent, mobile and interactive astronaut assistance system" that will join the International Space Station. These two events propel us into a debate over the risks created by the development of superintelligence that could eliminate jobs on a massive scale or, even worse, wipe the human species off the face of the planet – and raise the question of how to assess such a threat. To date, we have no experience of accidents or disasters due to faulty or malicious AI. However, the imaginations of artists and scientists are a treasure trove of material that tells the story of superintelligence freed from any human control.

It may have been overshadowed by Android P or a slightly terrifying AI phone call, but Google's new News app was one of the most important things the search giant announced at I/O this year. It relies completely on artificial intelligence to bring you breaking news, but the most fascinating -- and potentially most valuable -- addition is what Google calls Full Coverage. Fire up the News app on your Android phone and you'll see a tiny icon in the corner of certain stories -- tap that and you're taken to a dedicated page that surfaces related stories Google's AI deems trustworthy. Google's approach with the Full Coverage feature is fascinating for its consistency. Every person who clicks into a Full Coverage package gets the exact same blend of stories, every time.