The similarity of feature representations plays a pivotal role in the success of domain adaptation and generalization. Feature similarity includes both the invariance of marginal distributions and the closeness of conditional distributions given the desired response $y$ (e.g., class labels). Unfortunately, traditional methods always learn such features without fully taking into consideration the information in $y$, which in turn may lead to a mismatch of the conditional distributions or the mix-up of discriminative structures underlying data distributions. In this work, we introduce the recently proposed von Neumann conditional divergence to improve the transferability across multiple domains. We show that this new divergence is differentiable and eligible to easily quantify the functional dependence between features and $y$. Given multiple source tasks, we integrate this divergence to capture discriminative information in $y$ and design novel learning objectives assuming those source tasks are observed either simultaneously or sequentially. In both scenarios, we obtain favorable performance against state-of-the-art methods in terms of smaller generalization error on new tasks and less catastrophic forgetting on source tasks (in the sequential setup).

The ability to plan into the future while utilizing only raw high-dimensional observations, such as images, can provide autonomous agents with broad capabilities. Visual model-based reinforcement learning (RL) methods that plan future actions directly have shown impressive results on tasks that require only short-horizon reasoning, however, these methods struggle on temporally extended tasks. We argue that it is easier to solve long-horizon tasks by planning sequences of states rather than just actions, as the effects of actions greatly compound over time and are harder to optimize. To achieve this, we draw on the idea of collocation, which has shown good results on long-horizon tasks in optimal control literature, and adapt it to the image-based setting by utilizing learned latent state space models. The resulting latent collocation method (LatCo) optimizes trajectories of latent states, which improves over previously proposed shooting methods for visual model-based RL on tasks with sparse rewards and long-term goals. Videos and code at https://orybkin.github.io/latco/.

Conventional approaches to personalized dialogue generation typically require a large corpus, as well as predefined persona information. However, in a real-world setting, neither a large corpus of training data nor persona information are readily available. To address these practical limitations, we propose a novel multi-task meta-learning approach which involves training a model to adapt to new personas without relying on a large corpus, or on any predefined persona information. Instead, the model is tasked with generating personalized responses based on only the dialogue context. Unlike prior work, our approach leverages on the provided persona information only during training via the introduction of an auxiliary persona reconstruction task. In this paper, we introduce 2 frameworks that adopt the proposed multi-task meta-learning approach: the Multi-Task Meta-Learning (MTML) framework, and the Alternating Multi-Task Meta-Learning (AMTML) framework. Experimental results show that utilizing MTML and AMTML results in dialogue responses with greater persona consistency.

Two physicists, from EPFL and Columbia University, have introduced an approach for simulating the quantum approximate optimization algorithm using a traditional computer. Instead of running the algorithm on advanced quantum processors, the new approach uses a classical machine-learning algorithm that closely mimics the behavior of near-term quantum computers. In a paper published in Nature Quantum Information, EPFL professor Giuseppe Carleo and Matija Medvidović, a graduate student at Columbia University and at the Flatiron Institute in New York, have found a way to execute a complex quantum computing algorithm on traditional computers instead of quantum ones. The specific "quantum software" they are considering is known as Quantum Approximate Optimization Algorithm (QAOA) and is used to solve classical optimization problems in mathematics; it's essentially a way of picking the best solution to a problem out of a set of possible solutions. "There is a lot of interest in understanding what problems can be solved efficiently by a quantum computer, and QAOA is one of the more prominent candidates," says Carleo.

Our mission at Duolingo is to develop the best education in the world and make it universally available. But we've got more left to do -- and that's where you come in! Duolingo is the most popular language-learning application in the world, with over 500 million users and over half a billion exercises completed daily. Beyond our core learning product, we have also entered into literacy with Duolingo ABC and English proficiency testing with the Duolingo English Test. We are passionate about educating our users, making fact-based decisions, and finding innovative solutions to complex problems.

As we all know, Covid-19 struck Wuhan on New Year's Eve in 2019, and the city embraced total lockdown. Soon, the rest of China and the world followed. But what many don't know is Chinese education never went into lockdown. Within less than three months, Beijing Normal University (BNU) started a new semester offering more than 3,000 online courses. BNU could do so only because of China's comprehensive education technology (EdTech) drive in the preceding years, as a blog post at the Oxford Internet Institute elaborates.

Universities have been ramping up their data science education initiatives ever since 2012, when Tom Davenport and DJ Patil declared data scientist "the sexiest job of the 21st century" in the Harvard Business Review. According to the website Data Science Programs, there are more than 500 universities across the United States with data science degree programs. All told, there are more than 980 individual programs, with Master of Data Science being the most popular. This number has increased substantially in recent years, according to past numbers shared by this website. While the supply of data scientists emerging from universitites is up, strong demand for data scientists at American companies continues to outstrip supply, according to Martial Hebert, the dean of the School of Computer Science at Carnegie Mellon University.

Levi Belnap wants to make this clear from the beginning--artificial intelligence will never replace human teachers, at least not in our lifetime. What he and his colleagues at Merlyn Mind believe is that AI can enhance teachers' work. On this episode of EdTech Today, Levi introduces us to his nascent company's offering and provides some insights on how they believe the classroom experience can be better for all involved. The company launched out of stealth mode last month to unveil the first digital assistant built specifically for education that empowers teachers to more naturally use the technology in their classrooms and simplify their work. The company also announces it has closed $29 million in funding to date, led by Learn Capital.

Thanks to ever-advancing technology, educators now have access to incredibly useful tools that are more effective than anything they've ever had access to before. Augmented Reality (AR) is just one of the forms of technology that teachers are now using in their classrooms, and it's truly making a huge difference in the way that they teach and the way that their students are learning. AR is highly motivating and engaging. It's been found to be an incredibly effective way to teach students about highly advanced technological processes, such as STEM and coding; plus, it makes the process of learning this information faster and more fun. Add to that the fact that students are able to retain the information they are presented with via augmented technology and it's easy to see how it's changing the shape of learning.

Artificial intelligence was almost exclusively the domain of academic research for decades. In the past ten years, however, machine learning (ML) techniques have finally achieved sufficient effectiveness and practicality for large-scale adoption in companies and institutions. This adoption, however, remains incipient. Most organizations are still in the early stages gaining proficiency in these technologies and growing them at enterprise scale. The potential for professionals in this area, therefore, is enormous, evinced by the steady increase in ML job openings and courses. Given the proliferation of ML jobs postings out there, what are the roles and positions in the field?