Retrieving relevant targets from an extremely large target set under computational limits is a common challenge for information retrieval and recommendation systems. Tree models, which formulate targets as leaves of a tree with trainable node-wise scorers, have attracted a lot of interests in tackling this challenge due to their logarithmic computational complexity in both training and testing. Tree-based deep models (TDMs) and probabilistic label trees (PLTs) are two representative kinds of them. Though achieving many practical successes, existing tree models suffer from the training-testing discrepancy, where the retrieval performance deterioration caused by beam search in testing is not considered in training. This leads to an intrinsic gap between the most relevant targets and those retrieved by beam search with even the optimally trained node-wise scorers. We take a first step towards understanding and analyzing this problem theoretically, and develop the concept of Bayes optimality under beam search and calibration under beam search as general analyzing tools for this purpose. Moreover, to eliminate the discrepancy, we propose a novel algorithm for learning optimal tree models under beam search. Experiments on both synthetic and real data verify the rationality of our theoretical analysis and demonstrate the superiority of our algorithm compared to state-of-the-art methods.

This article is part of Demystifying AI, a series of posts that (try to) disambiguate the jargon and myths surrounding AI. Our Couch Conferences bring together industry experts to discuss what's next Which 10 items sold the most in the past year? From punch cards to keyboards, mice, and touch screens, human-computer interfacing technologies have undergone major changes, and each change has made it easier to make use of computing resources and power. But never have those changes been more dramatic than in the past decade, the period in which artificial intelligence turned from a sci-fi myth to everyday reality. Thanks to the advent of machine learning, AI algorithms that learn by examples, we can talk to Alexa, Siri, Cortana, and Google Assistant, and they can talk back to us.

This week brought good sales on Apple and Amazon devices, as well as some intriguing gaming deals. The Apple Watch Series 5 dropped to $299 again after WWDC kicked off earlier this week and Amazon still has some of its Echo speakers on sale (including the handy Echo Dot with clock). You can grab some extra storage for your Nintendo Switch for less at Newegg and Steam's Summer Sale has just begun. Here are the best deals we found this week that you can still get today. The latest Apple Watch has dropped to its lowest price ever again at Amazon and Walmart.

AI has become an integral part of every tech company's pitch to consumers. Fail to hype up machine learning or neural networks when unveiling a new product, and you might as well be hawking hand-cranked calculators. This can lead to overpromising. But judging by its recent WWDC performance, Apple has adopted a smarter and quieter approach. Sprinkled throughout Apple's announcements about iOS, iPadOS, and macOS were a number of features and updates that have machine learning at their heart.

The updated Amazon Alexa Plus is seen above. When it comes to making your home smart, you need the right gear. Amazon's Alexa resides in millions of households and make sure you get the most out of yours. Tap or click for 13 Alexa skills you'll use over and over again. While you're adjusting settings, it's not just snooping Big Tech companies you need to guard against.

Virtual assistants turn 16 this year and you don't have to look too hard - or speak too loudly - to find them. In fact, there will be around 8 billion voice-based devices by 2023 - more than the world's population today. From Amazon's Echo and Google's Assistant to Apple's Siri, Samsung's Bixby and Microsoft's Cortana, billions of people around the world are using their voices every day to schedule appointments, get directions, play music or get answers quickly-- all things that once required us to tediously type or write. Even Twitter recently announced that users can now audio tweet their inner musings. And yet, despite widespread adoption of voice-based devices in our personal lives, applications based on voice are nowhere as pervasive in our professional lives as they are in our homes. One could argue that consumer technology leads the way in changing human behavior and that the consumerization of the enterprise is always driven first by an expectation that work tools should be equally as convenient as personal technology solutions.

For example, NCF [8] takes user-item implicit feedback data as inputs for the rating prediction task; and DeepFM [6] leverages both numerical and categorical data for the CTR prediction task. However, high degree of specialization comes at the expense of model adaptability and tuning complexity. As recommendation tasks evolve over time and additional types of data are collected, the originally apt model can either become obsolete or require tremendous tuning efforts. So far, several pipelines for recommender systems, e.g., OpenRec [16] and SMORe [4], tried to address the adaptability issue via providing modular base blocks that can be selected according to the context of recommendation. Nevertheless, both determining the blocks to use and tuning the model parameters are not straightforward when facing new data and changing tasks. In order to bridge the gap, we present AutoRec, which aims to provide an end-to-end solution to automate model selection and hyperparameter tuning. While many AutoML libraries, such as Auto-Sklearn [5] and TPOT [12] have shown promising results in general-purpose machine learning tasks (e.g., regression and hyperparameter tuning) and

They might apply to one of these futuristic job ads one day. Fifteen years ago, people would have looked at you sideways if you told them you were a data scientist, driverless car engineer, or drone operator. It's hard to believe, but in 2006 those industries didn't really exist. By 2030, automation is expected to hit a midpoint, "something like 16 percent of occupations would have been automated--and there would be impact and dislocation as a result of these technologies." Artificial Intelligence, spatial computing (augmented and virtual reality), brain-computer interfaces, are all set to substitute labor or complement it in some way.

The digital revolution has been increasingly enhancing the impact of marketing on management. Digital marketing is quite a broad term encompassing social media marketing, e-mail marketing, internet marketing, and search engine marketing, among others. While e-mails, social media, websites, and e-commerce, are tools that marketers use to promote services and products to their target audience, Artificial Intelligence (AI) is now intrinsically being used for digital marketing to map the right kind of promotional tools for specific market offerings. AI and Customer Value A combination of AI and digital marketing has been adopted extensively across sectors such as healthcare, financial services, retail, automobiles, education, and entertainment. As a case in point, the ability to track data makes it possible for retailers to develop strategies for a better understanding of consumer behaviour by keeping the user's purchase life-cycle at the center.

Technology has made it easier for bright ideas to come into fruition, what with the available online tools and software that you can download and use even in the comfort of home. Selling that idea, however, is not as easy, and this is where some startups fall short. There are many considerations after transforming a great idea into something tangible for your target market. Not least of which is how a business can deliver the right product or service to the right market at the right time. This is where a go-to-market strategy comes into play.