On Tuesday, OpenAI announced that Elon Musk, one of the non-profit AI research company's founding members and foremost benefactors, would be vacating his position on the OpenAI board of directors. Musk helped craft OpenAI's vision and financed much of the nonprofit's growth. Elon Musk will depart the OpenAI Board but will continue to donate and advise the organization. As Tesla continues to become more focused on AI, this will eliminate a potential future conflict for Elon. The OpenAI board of directors now consists of Greg Brockman, Ilya Sutskever, Holden Karnofsky, and Sam Altman, with whom Musk co-founded the venture.

A retinal fundus image is a photograph of the back of the eye taken through the pupil. For more than 100 years these images have been used for detecting eye disease. Now Google has introduced a surprising new use for retinal images: combined with artificial intelligence, they can also predict a patient's risk of heart attack or stoke. Research arm Google Brain today published a paper in the journal Nature Biomedical Engineering which demonstrates how deep learning models can use retinal images to detect a patient's age, gender, smoking status and systolic blood pressure; calculate cardiovascular risk factors; and predict the risk of major adverse cardiac events occurring over the next five years. A problem with today's mainstream cardiovascular risk calculators such as the Pooled Cohort Equations, Framingham, and Systematic Coronary Risk Evaluation is that they require the input of multiple features such as blood pressure, body mass index, glucose and cholesterol levels, etc. to generate a disease risk result. A study by the American College of Cardiology's Practice Innovation And Clinical Excellence Program concluded that the data required to calculate 10-year risk was available for less than 30% of patients.

Artificial intelligence (AI) has the potential to revolutionize disease diagnosis and management by performing classification difficult for human experts and by rapidly reviewing immense amounts of images. Despite its potential, clinical interpretability and feasible preparation of AI remains challenging. The traditional algorithmic approach to image analysis for classification previously relied on (1) handcrafted object segmentation, followed by (2) identification of each segmented object using statistical classifiers or shallow neural computational machine-learning classifiers designed specifically for each class of objects, and finally (3) classification of the image (Goldbaum et al., 1996xSee all ReferencesGoldbaum et al., 1996). Creating and refining multiple classifiers required many skilled people and much time and was computationally expensive (Chaudhuri et al., 1989xDetection of blood vessels in retinal images using two-dimensional matched filters. The development of convolutional neural network layers has allowed for significant gains in the ability to classify images and detect objects in a picture (Krizhevsky et al., 2017xImageNet classification with deep convolutional neural networks. These are multiple processing layers to which image analysis filters, or convolutions, are applied.

In January 2018, both Microsoft's and Alibaba's deep-learning software did better than humans on reading and comprehension tests, opening the door to artificial intelligence–based customer service, medical diagnoses, and many other applications.

In January 2018, both Microsoft's and Alibaba's deep-learning software did better than humans on reading and comprehension tests, opening the door to artificial intelligence (AI) based customer service, medical inquiries, and many other applications.

Deep learning is a branch of machine learning which is all about learning of different levels of representation and abstraction. If you are working on a deep learning project then you need the smart deep learning platform to execute your code and to work over complex projects.

Note: If you are looking for a review paper, this blog post is also available as an article on arXiv. In Machine Learning (ML), we typically care about optimizing for a particular metric, whether this is a score on a certain benchmark or a business KPI. In order to do this, we generally train a single model or an ensemble of models to perform our desired task. While we can generally achieve acceptable performance this way, by being laser-focused on our single task, we ignore information that might help us do even better on the metric we care about. Specifically, this information comes from the training signals of related tasks. By sharing representations between related tasks, we can enable our model to generalize better on our original task. This approach is called Multi-Task Learning (MTL) and will be the topic of this blog post. Multi-task learning has been used successfully across all applications of machine learning, from natural language processing [1] and speech recognition [2] to computer vision [3] and drug discovery [4]. MTL comes in many guises: joint learning, learning to learn, and learning with auxiliary tasks are only some names that have been used to refer to it. Generally, as soon as you find yourself optimizing more than one loss function, you are effectively doing multi-task learning (in contrast to single-task learning).

Israeli TV to smartphone audio company Tunity has announced a $12 million Series A financing round from existing investors including former Morgan Stanley CEO John Mack and WeWork founder and CEO Adam Neumann, as well as new investors, including MGM Resorts International. Headquartered in New York and with its development center in Tel Aviv, Tunity has developed deep learning and computer vision based technology to produce an app that allows users to stream and hear live audio from muted televisions directly on their smartphones. Tunity will use the funding to complete the development of its data product, which is already attracting significant interest from television networks and advertisers, as well as to enhance its patented cloud-based Deep Learning and Computer Vision technology. Since its launch, the Tunity app has been downloaded more than 1.5 million times, further proving the company's value proposition to consumers through organic adoption and continued usage. After a user scans a nearby television screen, the Tunity app identifies the live video stream and its exact timing, syncing the audio with the user's mobile device.

Are you interested in some practical natural language processing resources? There are so many NLP resources available online, especially those relying on deep learning approaches, that sifting through to find the quality can be quite a task. But what if you've completed these, have already gained a foundation in NLP and want to move to some practical resources, or simply have an interest in other approaches, which may not necessarily be dependent on neural networks? This post (hopefully) will be helpful. This is the introductory natural language processing book, at least from the dual perspectives of practicality and the Python ecosystem.

When you get started with data science, you start simple. You go through simple projects like Loan Prediction problem or Big Mart Sales Prediction. These problems have structured data arranged neatly in a tabular format. In other words, you are spoon-fed the hardest part in data science pipeline. The datasets in real life are much more complex.