This paper considers online convex optimization (OCO) problems - the paramount framework for online learning algorithm design. The loss function of learning task in OCO setting is based on streaming data so that OCO is a powerful tool to model large scale applications such as online recommender systems. Meanwhile, real-world data are usually of extreme high-dimensional due to modern feature engineering techniques so that the quadratic regression is impractical. Factorization Machine as well as its variants are efficient models for capturing feature interactions with low-rank matrix model but they can't fulfill the OCO setting due to their non-convexity. In this paper, We propose a projective quadratic regression (PQR) model. First, it can capture the import second-order feature information. Second, it is a convex model, so the requirements of OCO are fulfilled and the global optimal solution can be achieved. Moreover, existing modern online optimization methods such as Online Gradient Descent (OGD) or Follow-The-Regularized-Leader (FTRL) can be applied directly. In addition, by choosing a proper hyper-parameter, we show that it has the same order of space and time complexity as the linear model and thus can handle high-dimensional data. Experimental results demonstrate the performance of the proposed PQR model in terms of accuracy and efficiency by comparing with the state-of-the-art methods.

Edy Portmann explains why it is important for schools to reinforce the scientist that lies within every child. He talks about intelligent learning systems and how they can be used to build collective intelligence, as well as to encourage students' creativity and help them learn to work together to solve problems. Sabine Gysi: In discussions of the digital transformation in education, skeptics often complain that reality is being pushed aside in favor of the digital. Does it make sense to look at the "real" world and the digital world as opposites? Edy Portmann: I've heard some teachers say that technological tools are "artificial."

There have been many words that have become almost extinct from our daily usage. But it seems that humanity next goal is to put the word "impossible" in the same section of deceased vocabularies. It is because with the progress that science and technology have made in this blink of an eye since human civilizations existed, almost nothing seems far-fetched these days. Science fiction might just be seen as a journal by people of the future to investigate our thought process. Just 150 years ago, the world was divided on the civil liberties of people dehumanized under slavery laws.

AI or Artificial Intelligence is fast emerging in the technology industry and is even seen taking center stage at several conferences. Its potential is showcased across industries with retail and manufacturing leading the way. Virtual assistants and chatbots are in-built into newer applications thus enabling easy customer interaction and product knowledge. The consumers can derive all that they require by simply logging into the suppliers' product page. Artificial Intelligence is also being integrated as an intelligence layer across the tech stacks by industry giants such as Microsoft, Google, and Salesforce.

Document recommendation systems for locating relevant literature have mostly relied on methods developed a decade ago. This is largely due to the lack of a large offline gold-standard benchmark of relevant documents that cover a variety of research fields such that newly developed literature search techniques can be compared, improved and translated into practice. To overcome this bottleneck, we have established the RElevant LIterature SearcH consortium consisting of more than 1500 scientists from 84 countries, who have collectively annotated the relevance of over 180 000 PubMed-listed articles with regard to their respective seed (input) article/s. The majority of annotations were contributed by highly experienced, original authors of the seed articles. The collected data cover 76% of all unique PubMed Medical Subject Headings descriptors. No systematic biases were observed across different experience levels, research fields or time spent on annotations.

Machine learning, a process in which artificial intelligence teaches itself to perform complex tasks, has boundless applications. But the risks are alarming. AI, for instance, could discriminate against hiring black people based on past trends when discrimination against them was rife. So, in order to avoid such undesirable behaviour, a team of computer scientists at Stanford University has developed a framework dubbed "Seldonian algorithms" after a character in the science-fiction novels of Isaac Asimov. Seldonian algorithms can easily be tweaked by end users--who may not be coding wizards--to pre-empt potential foul-ups.

At a packed hall in Lagos, a gathering of education and technology enthusiasts cheered for a milestone moment: the launch of Nigeria's first book on artificial intelligence for primary and secondary schools. The eight-chapter book illustrated with animations is written by Olubayo Adekanmbi, convener of Data Science Nigeria (DSN). His organisation has taken an active role in democratizing artificial intelligence application and research in Nigeria. With a suite of hands-on training programmes, toolkits and events, Data Science Nigeria aims to increase Nigeria's presence on the global AI map. "AI is a catalyst for good that creates new frontiers," Adekanmbi said, in his remarks at the launch.

Despite recent advances in architectures for mobile devices, deep learning computational requirements remains prohibitive for most embedded devices. To address that issue, we envision sharing the computational costs of inference between local devices and the cloud, taking advantage of the compression performed by the first layers of the networks to reduce communication costs. Inference in such distributed setting would allow new applications, but requires balancing a triple trade-off between computation cost, communication bandwidth, and model accuracy. We explore that trade-off by studying the compressibility of representations at different stages of MobileNetV2, showing those results agree with theoretical intuitions about deep learning, and that an optimal splitting layer for network can be found with a simple PCA-based compression scheme.

In this paper, we present a deep neural network (DNN) training approach called the "DeepMimic" training method. Enormous amounts of data are available nowadays for training usage. Yet, only a tiny portion of these data is manually labeled, whereas almost all of the data are unlabeled. The training approach presented utilizes, in a most simplified manner, the unlabeled data to the fullest, in order to achieve remarkable (classification) results. Our DeepMimic method uses a small portion of labeled data and a large amount of unlabeled data for the training process, as expected in a real-world scenario. It consists of a mentor model and a student model. Employing a mentor model trained on a small portion of the labeled data and then feeding it only with unlabeled data, we show how to obtain a (simplified) student model that reaches the same accuracy and loss as the mentor model, on the same test set, without using any of the original data labels in the training of the student model. Our experiments demonstrate that even on challenging classification tasks the student network architecture can be simplified significantly with a minor influence on the performance, i.e., we need not even know the original network architecture of the mentor. In addition, the time required for training the student model to reach the mentor's performance level is shorter, as a result of a simplified architecture and more available data. The proposed method highlights the disadvantages of regular supervised training and demonstrates the benefits of a less traditional training approach.

Imitation learning has gained immense popularity because of its high sample-efficiency. However, in real-world scenarios, where the trajectory distribution of most of the tasks dynamically shifts, model fitting on continuously aggregated data alone would be futile. In some cases, the distribution shifts, so much, that it is difficult for an agent to infer the new task. We propose a novel algorithm to generalize on any related task by leveraging prior knowledge on a set of specific tasks, which involves assigning importance weights to each past demonstration. We show experiments where the robot is trained from a diversity of environmental tasks and is also able to adapt to an unseen environment, using few-shot learning. We also developed a prototype robot system to test our approach on the task of visual navigation, and experimental results obtained were able to confirm these suppositions.