The term Artificial Intelligence (AI) was not coined yet and hardly anyone could follow what World famous scientist Alan Turing was talking about when he proposed a test which is now known as the "Turing Test" to answer the question "Can machines think?". When John McCarthy coined the term "Artificial Intelligence" around 1950, it was a dream to have machines that could actually think. Though we're not quite there yet, we are making rapid strides in the area of artificial intelligence and it is slowly making its presence felt in our lives. Let me illustrate where we will be heading in the coming 5 to 7 years. Imagine your present "you" in a self-driving car 5 years from now.

Over the last three years, there has been a significant shift towards the adoption of new and emerging technologies leading to a widening mix of advice, administration and investment practices among wealth managers. In the face of competitive market pressures, constant regulatory change and escalating data volumes, it is critical for wealth management firms to leverage technology and the underlying data creatively to improve service quality, personalise customer experiences and create platforms for smart processing. Wealth managers need to define and firmly establish digital operations as a capability within their businesses. One of the key drivers to the speed and adoption of digital operations is the maturity of the Robotic Automation solutions. There are two types of solutions in this marketplace: Unassisted Automation and Assisted Automation.

According to a memo obtained by sUAS News, the US Army will stop using DJI drones, effective immediately. "Due to increased awareness of cyber vulnerabilities associated with DJI products, it is directed that the US Army halt use of all DJI products," said the memo, which listed flight computers, cameras, radios, batteries, speed controllers, GPS units, handheld control stations and any device with DJI software applications installed on it as products that must cease being used. According to the document, the Army Aviation Engineering Directorate has issued over 300 Airworthiness Releases for DJI products. "Cease all use, uninstall all DJI applications, remove all batteries/storage media from devices, and secure equipment for follow on direction," the memo continued. The memo cites a report from the Army Research Laboratory and a memo from the US Navy, both compiled in May, that reference operational risks and vulnerabilities with DJI products.

This paper studies the landscape of empirical risk of deep neural networks by theoretically analyzing its convergence behavior to the population risk as well as its stationary points and properties. For an $l$-layer linear neural network, we prove its empirical risk uniformly converges to its population risk at the rate of $\mathcal{O}(r^{2l}\sqrt{d\log(l)}/\sqrt{n})$ with training sample size of $n$, the total weight dimension of $d$ and the magnitude bound $r$ of weight of each layer. We then derive the stability and generalization bounds for the empirical risk based on this result. Besides, we establish the uniform convergence of gradient of the empirical risk to its population counterpart. We prove the one-to-one correspondence of the non-degenerate stationary points between the empirical and population risks with convergence guarantees, which describes the landscape of deep neural networks. In addition, we analyze these properties for deep nonlinear neural networks with sigmoid activation functions. We prove similar results for convergence behavior of their empirical risks as well as the gradients and analyze properties of their non-degenerate stationary points. To our best knowledge, this work is the first one theoretically characterizing landscapes of deep learning algorithms. Besides, our results provide the sample complexity of training a good deep neural network. We also provide theoretical understanding on how the neural network depth $l$, the layer width, the network size $d$ and parameter magnitude determine the neural network landscapes.

Noname manuscript No. (will be inserted by the editor) Abstract Label embedding (LE) is an important family of multi-label classification algorithms that digest the label information jointly for better performance. Different real-world applications evaluate performance by different cost functions of interest. Current LE algorithms often aim to optimize one specific cost function, but they can suffer from bad performance with respect to other cost functions. In this paper, we resolve the performance issue by proposing a novel cost-sensitive LE algorithm that takes the cost function of interest into account. The proposed algorithm, cost-sensitive label embedding with multidimensional scaling (CLEMS), approximates the cost information with the distances of the embedded vectors by using the classic multidimensional scaling approach for manifold learning. CLEMS is able to deal with both symmetric and asymmetric cost functions, and effectively makes cost-sensitive decisions by nearest-neighbor decoding within the embedded vectors. We derive theoretical results that justify how CLEMS achieves the desired cost-sensitivity. Furthermore, extensive experimental results demonstrate that CLEMS is significantly better than a wide spectrum of existing LE algorithms and state-of-the-art cost-sensitive algorithms across different cost functions. Keywords Multi-label classification, Cost-sensitive, Label embedding 1 Introduction The multi-label classification problem (MLC), which allows multiple labels to be associated with each example, is an extension of the multi-class classification problem.

Connections between relations in relation extraction, which we call class ties, are common. In distantly supervised scenario, one entity tuple may have multiple relation facts. Exploiting class ties between relations of one entity tuple will be promising for distantly supervised relation extraction. However, previous models are not effective or ignore to model this property. In this work, to effectively leverage class ties, we propose to make joint relation extraction with a unified model that integrates convolutional neural network (CNN) with a general pairwise ranking framework, in which three novel ranking loss functions are introduced. Additionally, an effective method is presented to relieve the severe class imbalance problem from NR (not relation) for model training. Experiments on a widely used dataset show that leveraging class ties will enhance extraction and demonstrate the effectiveness of our model to learn class ties. Our model outperforms the baselines significantly, achieving state-of-the-art performance.

It was a return to the source for RoboCup 2017, which took place last week in Nagoya Japan, 20 years after its launch in the same city. Bigger than ever, the competition brought together roboticists from around the world. Originally focussed on robot football matches, RoboCup has expanded to include leagues for rescue robots, industrial robots, and robots in the home. Kids are also part of the fun, competing in their own matches and creative shows. You can watch video introductions of all of the leagues here, or watch a quick summary below.

Recently, IoT seems emerging as a solution vehicle for data analytics on real-world scenarios from setting a room temperature setting to predicting a component failure of an aircraft. Compared with developing an application or deploying a cloud service, is an IoT solution unique? How does a typical IoT solution architecture consist? And what are the essential components and how are they relevant to each other? How does the security play out?

A drone flies May 11, 2017, in the showroom of the DJI headquarters in Shenzhen, China. A Chinese company that is the world's largest drone manufacturer said Friday it is "surprised and disappointed" by reports the U.S. Army has halted use of its remote-controlled aircraft because of cyber vulnerabilities. An Army memo Wednesday, obtained by sUASnews.com The memo from Lt. Gen. Joseph Anderson, the deputy chief of staff, cited possible threats from any DJI electrical components, software, cameras, radios, GPS units or handheld controllers, the publications reported. It ordered U.S. Army personnel to uninstall all DJI applications and remove all batteries and media storage devices.

Video Friday is your weekly selection of awesome robotics videos, collected by your Automaton bloggers. We'll also be posting a weekly calendar of upcoming robotics events for the next two months; here's what we have so far (send us your events!): Let us know if you have suggestions for next week, and enjoy today's videos. It's not always clear whether Boston Dynamics' robots are operating autonomously or being controlled by a human, but it's definitely clear in this video of a presentation and demo by Marc Raibert: At the end, the human driving Atlas accidentally walks the robot off the stage. We developed a mobile robot that rolls by active deformation of the soft outer shell.