In this paper, we investigate the geometric structure of activation spaces of fully connected layers in neural networks and then show applications of this study. We propose an efficient approximation algorithm to characterize the convex hull of massive points in high dimensional space. Based on this new algorithm, four common geometric properties shared by the activation spaces are concluded, which gives a rather clear description of the activation spaces. We then propose an alternative classification method grounding on the geometric structure description, which works better than neural networks alone. Surprisingly, this data classification method can be an indicator of overfitting in neural networks. We believe our work reveals several critical intrinsic properties of modern neural networks and further gives a new metric for evaluating them.

Code-switching, the alternation of languages within a conversation or utterance, is a common communicative phenomenon that occurs in multilingual communities across the world. This survey reviews computational approaches for code-switched Speech and Natural Language Processing. We motivate why processing code-switched text and speech is essential for building intelligent agents and systems that interact with users in multilingual communities. As code-switching data and resources are scarce, we list what is available in various code-switched language pairs with the language processing tasks they can be used for. We review code-switching research in various Speech and NLP applications, including language processing tools and end-to-end systems. We conclude with future directions and open problems in the field.

In this paper, we consider high-dimensional nonconvex square-root-loss regression problems and introduce a proximal majorization-minimization (PMM) algorithm for these problems. Our key idea for making the proposed PMM to be efficient is to develop a sparse semismooth Newton method to solve the corresponding subproblems. By using the Kurdyka-{\L}ojasiewicz property exhibited in the underlining problems, we prove that the PMM algorithm converges to a d-stationary point. We also analyze the oracle property of the initial subproblem used in our algorithm. Extensive numerical experiments are presented to demonstrate the high efficiency of the proposed PMM algorithm.

Prognostics or Remaining Useful Life (RUL) Estimation from multi-sensor time series data is useful to enable condition-based maintenance and ensure high operational availability of equipment. We propose a novel deep learning based approach for Prognostics with Uncertainty Quantification that is useful in scenarios where: (i) access to labeled failure data is scarce due to rarity of failures (ii) future operational conditions are unobserved and (iii) inherent noise is present in the sensor readings. All three scenarios mentioned are unavoidable sources of uncertainty in the RUL estimation process often resulting in unreliable RUL estimates. To address (i), we formulate RUL estimation as an Ordinal Regression (OR) problem, and propose LSTM-OR: deep Long Short Term Memory (LSTM) network based approach to learn the OR function. We show that LSTM-OR naturally allows for incorporation of censored operational instances in training along with the failed instances, leading to more robust learning. To address (ii), we propose a simple yet effective approach to quantify predictive uncertainty in the RUL estimation models by training an ensemble of LSTM-OR models. Through empirical evaluation on C-MAPSS turbofan engine benchmark datasets, we demonstrate that LSTM-OR is significantly better than the commonly used deep metric regression based approaches for RUL estimation, especially when failed training instances are scarce. Further, our uncertainty quantification approach yields high quality predictive uncertainty estimates while also leading to improved RUL estimates compared to single best LSTM-OR models.

In this work, we present a method for node embedding in temporal graphs. We propose an algorithm that learns the evolution of a temporal graph's nodes and edges over time and incorporates this dynamics in a temporal node embedding framework for different graph prediction tasks. We present a joint loss function that creates a temporal embedding of a node by learning to combine its historical temporal embeddings, such that it optimizes per given task (e.g., link prediction). The algorithm is initialized using static node embeddings, which are then aligned over the representations of a node at different time points, and eventually adapted for the given task in a joint optimization. We evaluate the effectiveness of our approach over a variety of temporal graphs for the two fundamental tasks of temporal link prediction and multi-label node classification, comparing to competitive baselines and algorithmic alternatives. Our algorithm shows performance improvements across many of the datasets and baselines and is found particularly effective for graphs that are less cohesive, with a lower clustering coefficient.

Epistemic planning --- planning with knowledge and belief --- is essential in many multi-agent and human-agent interaction domains. Most state-of-the-art epistemic planners solve this problem by compiling to propositional classical planning, for example, generating all possible knowledge atoms, or compiling epistemic formula to normal forms. However, these methods become computationally infeasible as problems grow. In this paper, we decompose epistemic planning by delegating reasoning about epistemic formula to an external solver. We do this by modelling the problem using \emph{functional STRIPS}, which is more expressive than standard STRIPS and supports the use of external, black-box functions within action models. Exploiting recent work that demonstrates the relationship between what an agent `sees' and what it knows, we allow modellers to provide new implementations of externals functions. These define what agents see in their environment, allowing new epistemic logics to be defined without changing the planner. As a result, it increases the capability and flexibility of the epistemic model itself, and avoids the exponential pre-compilation step. We ran evaluations on well-known epistemic planning benchmarks to compare with an existing state-of-the-art planner, and on new scenarios based on different external functions. The results show that our planner scales significantly better than the state-of-the-art planner against which we compared, and can express problems more succinctly.

Broadband customers who are having internet problems are about to start getting refunds – without even having to ask. At the moment, only about one in seven people who have internet or landline problems such as repairs, installations or missed engineer appointments are given any kind of compensation from the companies responsible, according to regulator Ofcom. Even if they do, the amounts are usually small. But now customers will find themselves being given those refunds automatically, for any kind of broadband problems, Ofcom said. We'll tell you what's true.

No, this isn't an April Fool's joke: A Russian defense contractor has patented a drone that uses a shotgun to blast other drones out of the sky. It comes from Almaz Antey, a Russian defense contractor that manufactures the S-400 Triumf surface-to-air missile that caused a rift between Turkey and the US. The tail-sitting drone takes off on the spot but flies like an airplane for greater efficiency, giving it a 40-minute range while packing a fully-automatic Vepr-12 shotgun with a 10-round magazine. The drone was built by the "Student Design Bureau of Aviation Modeling" at the Moscow Aviation Institute for Almaz Antey. It's of a similar type used by mining companies, farmers and others to survey pipelines and other installations. A visor-wearing operator uses a live video link to fly the drone and aim the weapon, which is tucked into the nose of the aircraft.

Speaking at MIT Technology Review's EmTech Digital event, Microsoft's Vice President of AI & Research, Harry Shum, drew attention to the risks associated with AI as it becomes more creative in the future. In particular, Shum called on tech companies to "engineer responsibility into the very fabric of the technology. As MIT's Technology Review points out, we've already seen some of the fallout from the tech industry failing to anticipate flaws in AI. One such flaw is AI's difficulty thus far with identifying faces with dark skin tones, something Microsoft has been working to improve. But AI is also being used by China in alarming ways for surveillance, while, more recently, an Uber self-driving car killed a pedestrian in early 2018. According to Shum, AI's challenges will only ramp up as it becomes more complex, adding the ability to produce art, maintain near-human-like conversations, and accurately read human emotions. These abilities will pave the way for AI to more easily create propaganda or misinformation to be spread online, including fake audio and video. Microsoft is working to take these challenges into account. The company has created an AI ethics committee and is working with others in the industry to address problems posed by AI. Shum also told MIT Technology Review that Microsoft plans to add an ethics review step to its audit list before products hit the market "one day very soon," joining other steps such as privacy, security, and accessibility. "We are working hard to get ahead of the challenges posed by AI creation," Shum told MIT Technology Review. "But these are hard problems that can't be solved with technology alone, so we really need the cooperation across academia and industry.

For unsupervised domain adaptation problems, the strategy of aligning the two domains in latent feature space through adversarial learning has achieved much progress in image classification, but usually fails in semantic segmentation tasks in which the latent representations are overcomplex. In this work, we equip the adversarial network with a "significance-aware information bottleneck (SIB)", to address the above problem. The new network structure, called SIBAN, enables a significance-aware feature purification before the adversarial adaptation, which eases the feature alignment and stabilizes the adversarial training course. In two domain adaptation tasks, i.e., GTA5 -> Cityscapes and SYNTHIA -> Cityscapes, we validate that the proposed method can yield leading results compared with other feature-space alternatives. Moreover, SIBAN can even match the state-of-the-art output-space methods in segmentation accuracy, while the latter are often considered to be better choices for domain adaptive segmentation task.