Subscribe to stay notified about new videos: http://3b1b.co/subscribe Typo correction: At 14:45, the last index on the bias vector is n, when it's supposed to in fact be a k. Thanks for the sharp eyes that caught that! For those who want to learn more, I highly recommend the book by Michael Nielsen introducing neural networks and deep learning: https://goo.gl/Zmczdy There are two neat things about this book.

Pyro is a universal probabilistic programming language (PPL) written in Python and supported by PyTorch on the backend. Pyro enables flexible and expressive deep probabilistic modeling, unifying the best of modern deep learning and Bayesian modeling. It was designed with these key principles: Universal: Pyro can represent any computable probability distribution. Minimal: Pyro is implemented with a small core of powerful, composable abstractions. Flexible: Pyro aims for automation when you want it, control when you need it.

Medical data is horrible to work with. In medical imaging, data stores (archives) operate on clinical assumptions. Unfortunately, this means that when you want to extract an image (say a frontal chest x-ray), you will often get a folder full of other images with no easy way to tell them apart. Depending on the manufacturer, you might end up with horizontally or vertically flipped images. They might have inverted pixel values. The question is, when dealing with a huge dataset (say, 50-100k images), how do you find these aberrations without having a doctor look at all of them?

A fully automated artificial intelligence (AI)-based multispectral absorbance imaging system effectively classified function and potency of induced pluripotent stem cell derived retinal pigment epithelial cells (iPSC-RPE) from patients with age-related macular degeneration (AMD). The finding from the system could be applied to assessing future cellular therapies, according to research presented at the 2018 ARVO annual meeting. The software, which uses convolutional neural network (CNN) deep learning algorithms, effectively evaluated release criterion for the iPSC-RPE cell-based therapy in a standard, reproducible, and cost-effective fashion. The AI-based analysis was as specific and sensitive as traditional molecular and physiological assays, without the need for human intervention. "Cells can be classified with high accuracy using nothing but absorbance images," wrote lead investigator Nathan Hotaling and colleagues from the National Institutes of Health in their poster.

Suggestion: require a salary indication from job postings, or at least promote the ones that have one.

Google cofounder Sergey Brin speaks during a press conference after the third game of the Google DeepMind Challenge Match against Google-developed supercomputer AlphaGo at a hotel in Seoul on March 12, 2016. A new artificial intelligence ethics (AI) board was announced this week by Axon -- the US company behind the taser weapon -- but the AI ethics board many people still want to know about remains shrouded in mystery. Google quietly set up an AI ethics board in 2014 following the £400 million acquisition of a London AI lab called DeepMind, which hopes to one day build machines with human-level intelligence that will have a profound impact on the society we live in. Who sits on that board, how often that board meets, or what that board discusses, has remained a closely guarded company secret, despite DeepMind cofounder Mustafa Suleyman (who lobbied for the creation of the board) saying in 2016 that Google will publicise the names of those on it. This week, Axon, a US company that develops body cameras for police officers and weapons for the law enforcement market, demonstrated the kind of transparency that Google should aspire towards when it announced an AI ethics board to "help guide the development of Axon's AI-powered devices and services".

Existing methods for interactive image retrieval have demonstrated the merit of integrating user feedback, improving retrieval results. However, most current systems rely on restricted forms of user feedback, such as binary relevance responses, or feedback based on a fixed set of relative attributes, which limits their impact. In this paper, we introduce a new approach to interactive image search that enables users to provide feedback via natural language, allowing for more natural and effective interaction. We formulate the task of dialog-based interactive image retrieval as a reinforcement learning problem, and reward the dialog system for improving the rank of the target image during each dialog turn. To avoid the cumbersome and costly process of collecting human-machine conversations as the dialog system learns, we train our system with a user simulator, which is itself trained to describe the differences between target and candidate images. The efficacy of our approach is demonstrated in a footwear retrieval application. Extensive experiments on both simulated and real-world data show that 1) our proposed learning framework achieves better accuracy than other supervised and reinforcement learning baselines and 2) user feedback based on natural language rather than pre-specified attributes leads to more effective retrieval results, and a more natural and expressive communication interface.

In the last years, deep learning algorithms have highly revolutionized several areas including speech, image and natural language processing. The specific field of Machine Translation (MT) has not remained invariant. Integration of deep learning in MT varies from re-modeling existing features into standard statistical systems to the development of a new architecture. Among the different neural networks, research works use feedforward neural networks, recurrent neural networks and the encoder-decoder schema. These architectures are able to tackle challenges as having low-resources or morphology variations. This manuscript focuses on describing how these neural networks have been integrated to enhance different aspects and models from statistical MT, including language modeling, word alignment, translation, reordering, and rescoring. Then, we report the new neural MT approach together with a description of the foundational related works and recent approaches on using subword, characters and training with multilingual languages, among others. Finally, we include an analysis of the corresponding challenges and future work in using deep learning in MT.

In an influential paper (“Rule Learning by Seven-Month-Old Infants”), Marcus, Vijayan, Rao and Vishton claimed that connectionist models cannot account for human success at learning tasks that involved generalization of abstract knowledge such as grammatical rules. This claim triggered a heated debate, centered mostly around variants of the Simple Recurrent Network model. In our work, we revisit this unresolved debate and analyze the underlying issues from a different perspective. We argue that, in order to simulate human-like learning of grammatical rules, a neural network model should not be used as a tabula rasa, but rather, the initial wiring of the neural connections and the experience acquired prior to the actual task should be incorporated into the model. We present two methods that aim to provide such initial state: a manipulation of the initial connections of the network in a cognitively plausible manner (concretely, by implementing a “delay-line” memory), and a pre-training algorithm that incrementally challenges the network with novel stimuli. We implement such techniques in an Echo State Network (ESN), and we show that only when combining both techniques the ESN is able to learn truly general identity rules. Finally, we discuss the relation between these cognitively motivated techniques and recent advances in Deep Learning.

We investigate how neural networks can learn and process languages with hierarchical, compositional semantics. To this end, we define the artificial task of processing nested arithmetic expressions, and study whether different types of neural networks can learn to compute their meaning. We find that recursive neural networks can implement a generalising solution to this problem, and we visualise this solution by breaking it up in three steps: project, sum and squash. As a next step, we investigate recurrent neural networks, and show that a gated recurrent unit, that processes its input incrementally, also performs very well on this task: the network learns to predict the outcome of the arithmetic expressions with high accuracy, although performance deteriorates somewhat with increasing length. To develop an understanding of what the recurrent network encodes, visualisation techniques alone do not suffice. Therefore, we develop an approach where we formulate and test multiple hypotheses on the information encoded and processed by the network. For each hypothesis, we derive predictions about features of the hidden state representations at each time step, and train 'diagnostic classifiers' to test those predictions. Our results indicate that the networks follow a strategy similar to our hypothesised 'cumulative strategy', which explains the high accuracy of the network on novel expressions, the generalisation to longer expressions than seen in training, and the mild deterioration with increasing length. This in turn shows that diagnostic classifiers can be a useful technique for opening up the black box of neural networks. We argue that diagnostic classification, unlike most visualisation techniques, does scale up from small networks in a toy domain, to larger and deeper recurrent networks dealing with real-life data, and may therefore contribute to a better understanding of the internal dynamics of current state-of-the-art models in natural language processing.