If you are practitioner interested in deep learning for NLP, you may have different goals and requirements from the material. For example, you may want to focus on the methods and applications rather than the foundational theory. The course is comprised of 13 lectures, although the first and second lectures are both split into two parts. The complete lecture breakdown is provided below. The GitHub repository for the course provides links to slides, flash videos and reading for each lecture. I would recommend watching the videos via this unofficial YouTube playlist. Below is a course overview slide taken from the first lecture.

Every face does not tell a story; it tells thousands of them. Over evolutionary time, the human brain has become an exceptional reader of the human face--computerlike, we like to think. A viewer instinctively knows the difference between a real smile and a fake one. In July, a Canadian study reported that college students can reliably tell if people are richer or poorer than average simply by looking at their expressionless faces. Scotland Yard employs a team of "super-recognizers" who can, from a pixelated photo, identify a suspect they may have seen briefly years earlier or come across in a mug shot.

Inc. (NASDAQ:AMZN) has become one of the leaders in the emerging field of artificial intelligence (AI). Many of the ways the company uses the technology go on behind the scenes for things such as improving search results, better product recommendations, and enhanced forecasting and inventory management. The more public face of Amazon's AI ambitions can be found in its Echo family of smart speakers powered by Alexa, its voice-activated digital assistant, which has become the leader in the burgeoning smart-speaker market. With that much riding on AI, it should come as no surprise that Amazon has announced plans to expand its AI research by building a new R&D center in Barcelona, Spain. Amazon is setting up AI labs around the globe in search of AI talent.

Christoffer O. Hernæs is chief digital officer of Skandiabanken, Norway's first pure internet bank and leading challenger bank. Artificial intelligence is one of the hottest subjects these days, and recent advances in technology make AI even closer to reality than most of us can imagine. The subject really got traction when Stephen Hawking, Elon Musk and more than 1,000 AI and robotics researchers signed an open letter issuing a warning regarding the use of AI in weapons development last year. The following month, BAE Systems unveiled Taranis, the most advanced autonomous UAV ever created; there are currently 40 countries working on the deployment of AI in weapons development. Those in the defense industry are not the only ones engaging in an arms race to create advanced AI. Tech giants Facebook, Google, Microsoft and IBM are all engaging in various AI-initiatives, as well as competing on developing digital personal assistants like Facebook's M, Cortana from Microsoft and Apple' Siri.

Software that carries out tasks by itself sounds like sci-fi, but increasingly, it's part of everyday life – from self-driving cars to intelligent homes. For now, most smart technology is still designed in the US. Germany has some catching up to do.

The vehicle to represent Knowledge Organization Systems (KOSs) in the environment of the Semantic Web and linked data is the Simple Knowledge Organization System (SKOS). SKOS provides a way to assign a URI to each concept, and this URI functions as a surrogate for the concept. This fact makes of main concern the need to clarify the URIs' ontological meaning. The aim of this study is to investigate the relation between the ontological substance of KOS concepts and concepts revealed through the grammatical and syntactic formalisms of natural language. For this purpose, we examined the dividableness of concepts in specific KOSs (i.e. a thesaurus, a subject headings system and a classification scheme) by applying Natural Language Processing (NLP) techniques (i.e. morphosyntactic analysis) to the lexical representations (i.e. RDF literals) of SKOS concepts. The results of the comparative analysis reveal that, despite the use of multi-word units, thesauri tend to represent concepts in a way that can hardly be further divided conceptually, while Subject Headings and Classification Schemes - to a certain extent - comprise terms that can be decomposed into more conceptual constituents. Consequently, SKOS concepts deriving from thesauri are more likely to represent atomic conceptual units and thus be more appropriate tools for inference and reasoning. Since identifiers represent the meaning of a concept, complex concepts are neither the most appropriate nor the most efficient way of modelling a KOS for the Semantic Web.

This paper argues that there has not been enough discussion in the field of applications of Gaussian Process for the fast moving consumer goods industry. Yet, this technique can be important as it e.g., can provide automatic feature relevance determination and the posterior mean can unlock insights on the data. Significant challenges are the large size and high dimensionality of commercial data at a point of sale. The study reviews approaches in the Gaussian Processes modeling for large data sets, evaluates their performance on commercial sales and shows value of this type of models as a decision-making tool for management.

Labeled Latent Dirichlet Allocation (LLDA) is an extension of the standard unsupervised Latent Dirichlet Allocation (LDA) algorithm, to address multi-label learning tasks. Previous work has shown it to perform in par with other state-of-the-art multi-label methods. Nonetheless, with increasing label sets sizes LLDA encounters scalability issues. In this work, we introduce Subset LLDA, a simple variant of the standard LLDA algorithm, that not only can effectively scale up to problems with hundreds of thousands of labels but also improves over the LLDA state-of-the-art. We conduct extensive experiments on eight data sets, with label sets sizes ranging from hundreds to hundreds of thousands, comparing our proposed algorithm with the previously proposed LLDA algorithms (Prior--LDA, Dep--LDA), as well as the state of the art in extreme multi-label classification. The results show a steady advantage of our method over the other LLDA algorithms and competitive results compared to the extreme multi-label classification algorithms.

We introduce Latent Gaussian Process Regression which is a latent variable extension allowing modelling of non-stationary multi-modal processes using GPs. The approach is built on extending the input space of a regression problem with a latent variable that is used to modulate the covariance function over the training data. We show how our approach can be used to model multi-modal and non-stationary processes. We exemplify the approach on a set of synthetic data and provide results on real data from motion capture and geostatistics.

Dimension reduction is the process of embedding high-dimensional data into a lower dimensional space to facilitate its analysis. In the Euclidean setting, one fundamental technique for dimension reduction is to apply a random linear map to the data. This dimension reduction procedure succeeds when it preserves certain geometric features of the set. The question is how large the embedding dimension must be to ensure that randomized dimension reduction succeeds with high probability. This paper studies a natural family of randomized dimension reduction maps and a large class of data sets. It proves that there is a phase transition in the success probability of the dimension reduction map as the embedding dimension increases. For a given data set, the location of the phase transition is the same for all maps in this family. Furthermore, each map has the same stability properties, as quantified through the restricted minimum singular value. These results can be viewed as new universality laws in high-dimensional stochastic geometry. Universality laws for randomized dimension reduction have many applications in applied mathematics, signal processing, and statistics. They yield design principles for numerical linear algebra algorithms, for compressed sensing measurement ensembles, and for random linear codes. Furthermore, these results have implications for the performance of statistical estimation methods under a large class of random experimental designs.