Here's the trick: By then, computers will possess emotions and personality. "When I talk about computers reaching human levels of intelligence, I'm not talking about logical intelligence," Kurzweil said at an event in New York on Monday night. "It is being funny, and expressing a loving sentiment... That is the cutting edge of human intelligence." The futurist spoke about AI and the future of technology with astrophysicist Neil deGrasse Tyson.

SoftBank Group will shortly launch a service that will help clients thwart cyberattacks by using artificial intelligence. The company has established a joint venture in Japan with Cybereason, a U.S. cybersecurity firm based in Boston, in which SoftBank has invested roughly 50 million. The Boston firm will provide the service in Japan through the venture, targeting manufacturers, financial institutions and government-related agencies that handle sensitive information. SoftBank will handle sales, pitching the service to Japanese prospects. The joint venture will provide Japanese-language customer support.

This paper presents a Bayesian optimization method with exponential convergence without the need of auxiliary optimization and without the delta-cover sampling. Most Bayesian optimization methods require auxiliary optimization: an additional non-convex global optimization problem, which can be time-consuming and hard to implement in practice. Also, the existing Bayesian optimization method with exponential convergence requires access to the delta-cover sampling, which was considered to be impractical. Our approach eliminates both requirements and achieves an exponential convergence rate.

Feature extraction has gained increasing attention in the field of machine learning, as in order to detect patterns, extract information, or predict future observations from big data, the urge of informative features is crucial. The process of extracting features is highly linked to dimensionality reduction as it implies the transformation of the data from a sparse high-dimensional space, to higher level meaningful abstractions. This dissertation employs Neural Networks for distributed paragraph representations, and Latent Dirichlet Allocation to capture higher level features of paragraph vectors. Although Neural Networks for distributed paragraph representations are considered the state of the art for extracting paragraph vectors, we show that a quick topic analysis model such as Latent Dirichlet Allocation can provide meaningful features too. We evaluate the two methods on the CMU Movie Summary Corpus, a collection of 25,203 movie plot summaries extracted from Wikipedia. Finally, for both approaches, we use K-Nearest Neighbors to discover similar movies, and plot the projected representations using T-Distributed Stochastic Neighbor Embedding to depict the context similarities. These similarities, expressed as movie distances, can be used for movies recommendation. The recommended movies of this approach are compared with the recommended movies from IMDB, which use a collaborative filtering recommendation approach, to show that our two models could constitute either an alternative or a supplementary recommendation approach.

This paper presents a novel latent variable recurrent neural network architecture for jointly modeling sequences of words and (possibly latent) discourse relations between adjacent sentences. A recurrent neural network generates individual words, thus reaping the benefits of discriminatively-trained vector representations. The discourse relations are represented with a latent variable, which can be predicted or marginalized, depending on the task. The resulting model can therefore employ a training objective that includes not only discourse relation classification, but also word prediction. As a result, it outperforms state-of- the-art alternatives for two tasks: implicit discourse relation classification in the Penn Discourse Treebank, and dialog act classification in the Switchboard corpus. Furthermore, by marginalizing over latent discourse relations at test time, we obtain a discourse informed language model, which improves over a strong LSTM baseline.

Similarity metrics are a core component of many information retrieval and machine learning systems. In this work we propose a method capable of learning a similarity metric from data equipped with a binary relation. By considering only the similarity constraints, and initially ignoring the features, we are able to learn target vectors for each instance using one of several appropriately designed loss functions. A regression model can then be constructed that maps novel feature vectors to the same target vector space, resulting in a feature extractor that computes vectors for which a predefined metric is a meaningful measure of similarity. We present results on both multiclass and multi-label classification datasets that demonstrate considerably faster convergence, as well as higher accuracy on the majority of the intrinsic evaluation tasks and all extrinsic evaluation tasks.

Are you struggling to find the right resources to guide you in your machine learning journey? Sign-up for the Machine Learning Resource Guide and you'll get a hand-picked list of the best resources for Machine Learning. In addition, you'll get 2-to-3 short, educational emails every week with actionable advice on how to further your study of Machine Learning. Yes, I want the best resources for Machine Learning (free)! I want to get started with Machine Learning Algorithms! Gentle introduction to 10 top machine learning algorithms and 12 tutorials to implement them step-by-step from scratch.

Salesforce has acquired artificial intelligence startup MetaMind. Terms of the deal were not disclosed. Based in Palo Alto, MetaMind -- which was backed by Salesforce CEO Marc Benioff -- touts its technology as being enterprise-grade AI. Its work has revolved mainly around deep learning in language and vision-based applications. "With MetaMind and Salesforce coming together, we'll be able to offer customers real AI solutions with breakthrough capabilities that further automate and personalize customer support, marketing automation, and many other business processes," MetaMind cofounder and chief executive Richard Socher wrote in a blog post.

A big shakeup happening in the world of deep learning, as Salesforce announced that it has acquired startup darling MetaMind. As part of the acquisition, MetaMind will shut down on May 4 for unpaid users and June 4 for paid users. Don't miss our biggest TNW Conference yet! With MetaMind and Salesforce coming together, we'll be able to offer customers real AI solutions with breakthrough capabilities that further automate and personalize customer support, marketing automation, and many other business processes. We'll extend Salesforce's data science capabilities by embedding deep learning within the Salesforce platform.

MetaMind, a Palo Alto-based AI startup founded in July 2014, is being acquired by Salesforce. According to a new post published at the company's website by CEO Richard Socher -- a Stanford PhD who studied machine learning, deep learning, natural language processing and computer vision -- Salesforce plans to use its technology to "further automate and personalize customer support, marketing automation, and many other business processes. Salesforce confirmed the deal but isn't disclosing financial details of the transaction or commenting on whether MetaMind's entire team will join its ranks. As a standalone company, MetaMind's general-purpose platform was designed to predict outcomes for language, vision and database tasks. As of the middle of last year, its technology could reportedly answer everything from specific queries about snippets of text to the sentiment of that text.