You immigrate to a new country that speaks a different language, and start work with some of the brightest engineers in the world. Now, you're leading teams of people who are 10 or 20 years older than you, working on one of the fastest growing internet companies of the last decade. You have two options: sink or swim. That's the position Simon Eskildsen found himself in early in his career. He left his home in Denmark after high school, and moved to Canada alone to take a pre-college gap year working at Shopify. When he started, Shopify had 150 employees supporting tens of thousands of merchants. Now, it has 5,000 employees and over a million merchants.

Computer scientists from Rice, supported by collaborators from Intel, will present their results today at the Austin Convention Center as a part of the machine learning systems conference MLSys. Many companies are investing heavily in GPUs and other specialized hardware to implement deep learning, a powerful form of artificial intelligence that's behind digital assistants like Alexa and Siri, facial recognition, product recommendation systems and other technologies. For example, Nvidia, the maker of the industry's gold-standard Tesla V100 Tensor Core GPUs, recently reported a 41% increase in its fourth quarter revenues compared with the previous year. Rice researchers created a cost-saving alternative to GPU, an algorithm called "sub-linear deep learning engine" (SLIDE) that uses general purpose central processing units (CPUs) without specialized acceleration hardware. "Our tests show that SLIDE is the first smart algorithmic implementation of deep learning on CPU that can outperform GPU hardware acceleration on industry-scale recommendation datasets with large fully connected architectures," said Anshumali Shrivastava, an assistant professor in Rice's Brown School of Engineering who invented SLIDE with graduate students Beidi Chen and Tharun Medini.

We revisit the problem of online learning with sleeping experts/bandits: in each time step, only a subset of the actions are available for the algorithm to choose from (and learn about). The work of Kleinberg et al. [2010] showed that there exist no-regret algorithms which perform no worse than the best ranking of actions asymptotically. Unfortunately, achieving this regret bound appears computationally hard: Kanade and Steinke [2014] showed that achieving this no-regret performance is at least as hard as PAC-learning DNFs, a notoriously difficult problem. In the present work, we relax the original problem and study computationally efficient no-approximate-regret algorithms: such algorithms may exceed the optimal cost by a multiplicative constant in addition to the additive regret. We give an algorithm that provides a no-approximate-regret guarantee for the general sleeping expert/bandit problems. For several canonical special cases of the problem, we give algorithms with significantly better approximation ratios; these algorithms also illustrate different techniques for achieving no-approximate-regret guarantees.

We investigate the capacity control provided by dropout in various machine learning problems. First, we study dropout for matrix completion, where it induces a data-dependent regularizer that, in expectation, equals the weighted trace-norm of the product of the factors. In deep learning, we show that the data-dependent regularizer due to dropout directly controls the Rademacher complexity of the underlying class of deep neural networks. These developments enable us to give concrete generalization error bounds for the dropout algorithm in both matrix completion as well as training deep neural networks. We evaluate our theoretical findings on real-world datasets, including MovieLens, MNIST, and Fashion-MNIST.

Portfolio Selection is an important real-world financial task and has attracted extensive attention in artificial intelligence communities. This task, however, has two main difficulties: (i) the non-stationary price series and complex asset correlations make the learning of feature representation very hard; (ii) the practicality principle in financial markets requires controlling both transaction and risk costs. Most existing methods adopt handcraft features and/or consider no constraints for the costs, which may make them perform unsatisfactorily and fail to control both costs in practice. In this paper, we propose a cost-sensitive portfolio selection method with deep reinforcement learning. Specifically, a novel two-stream portfolio policy network is devised to extract both price series patterns and asset correlations, while a new cost-sensitive reward function is developed to maximize the accumulated return and constrain both costs via reinforcement learning. We theoretically analyze the near-optimality of the proposed reward, which shows that the growth rate of the policy regarding this reward function can approach the theoretical optimum. We also empirically evaluate the proposed method on real-world datasets. Promising results demonstrate the effectiveness and superiority of the proposed method in terms of profitability, cost-sensitivity and representation abilities.

NASA has equipped its Mars 2020 rover with everything it needs to explore the Red planet, except for a name – until now. Called Perseverance, the rover's title was picked from a'Name the Rover' essay contest that received 28,000 entries from children ranging from kindergartners to high school. The name was revealed on Thursday during a live streaming and was chosen by seventh grader Alex Mathers who's winning essay compared the rover to the human race. 'If you think about it, all of these names of past Mars rovers are qualities we possess as humans.' 'We are always curious, and seek opportunity. We have the spirit and insight to explore the Moon, Mars, and beyond. But, if rovers are to be the qualities of us as a race, we missed the most important thing.

Artificial Intelligence is coming for education. It's not going to replace college faculty or teaching as we know it. Instead, AI is going to give faculty superpowers, extending their reach and expanding their time. A good teacher is a role model, a sage, able to become what the student needs. Teaching is too personal, too human, to be turned over to AI.

Machine learning and neural networks are fast becoming pillars on which you can build intelligent applications. The course will begin by introducing you to Python and discussing using AI search algorithms. You will learn math-heavy topics, such as regression and classification, illustrated by Python examples. You will then progress on to advanced AI techniques and concepts, and work on real-life data sets to form decision trees and clusters. You will be introduced to neural networks, which is a powerful tool benefiting from Moore's law applied on 21st-century computing power.

Data Science is about explaining the past and predicting the future by means of data analysis. Data Science is a multi-disciplinary field which combines statistics, machine learning, artificial intelligence and database technology. This course provides the essential concepts and principles in data science. Students learns commonly used classification algorithms and how to use those algorithms to solve real world problems.

New York is the most talented city in the world, according to our Global Talent Competitiveness Index 2020. London took second place and Singapore third. What makes these places so attractive to workers and are they ready for the Future of Work? A different workforce will be needed in the future as automation, machine learning and AI tech start to reach all corners of the employment market. As well as a places ability to grow, attract and retain talent, this year's GTCI report took into account a city's affinity for new tech to measure how prepared they are for the changing digital work environment.