Japanese technology company SoftBank Group Corp. has reported a 98 percent drop in its April-June profit on losses stemming from investments in the Chinese e-commerce company Alibaba. SoftBank said Monday its quarterly net profit was 5.5 billion yen ($50 million), down from 254 billion yen the previous year. Quarterly sales added 3 percent to 2.19 trillion yen ($20 billion). The Tokyo-based company's operating profit, which highlights core operations, logged a 50 percent increase year-on-year as its U.S. mobile carrier Sprint boosted profitability. Softbank, which sells the Pepper robot, did not give an annual forecast, which is not unusual for the company.

More than 50 million Americans suffer from sleep disorders, and diseases including Parkinson's and Alzheimer's can also disrupt sleep. To make it easier to diagnose and study sleep problems, researchers at MIT and Massachusetts General Hospital have devised a new way to monitor sleep stages without sensors attached to the body. Their device uses an advanced artificial intelligence algorithm to analyze the radio signals around the person and translate those measurements into sleep stages: light, deep, or rapid eye movement (REM). Recent advances in artificial intelligence have made it possible to train computer algorithms known as deep neural networks to extract and analyze information from complex datasets, such as the radio signals obtained from the researchers' sensor.

This unique, innovative multidisciplinary approach opens new insights and advances our understanding of the neurobiology of schizophrenia, which may help to improve the treatment and management of the disease,

Honda Motor Co.'s go-it-alone strategy looks increasingly risky as an alliance-happy auto industry scales up to cope with the soaring investment needed for self-driving vehicles, electrified power trains, connected-car technologies and artificial intelligence. In the latest industry tieup, Toyota Motor Corp. and Mazda Motor Corp. on Friday announced plans to jointly invest in a $1.6 billion assembly plant in the U.S. with the capacity to produce 300,000 Toyota Corollas and a new Mazda crossover model starting 2021. Toyota will acquire about 5 percent in Mazda, which will hold a 0.25 percent in the bigger automaker. The duo will also combine engineering efforts on electric vehicle development and car-networking know-how. Honda President and CEO Takahiro Hachigo, like his predecessors, isn't a fan of big alliances that involve equity stakes.

Approximate Bayesian computation (ABC) is now an established technique for statistical inference used in cases where the likelihood function is computationally expensive or not available. It relies on the use of a model that is specified in the form of a simulator, and approximates the likelihood at a parameter $\theta$ by simulating auxiliary data sets $x$ and evaluating the distance of $x$ from the true data $y$. However, ABC is not computationally feasible in cases where using the simulator for each $\theta$ is very expensive. This paper investigates this situation in cases where a cheap, but approximate, simulator is available. The approach is to employ delayed acceptance Markov chain Monte Carlo (MCMC) within an ABC sequential Monte Carlo (SMC) sampler in order to, in a first stage of the kernel, use the cheap simulator to rule out parts of the parameter space that are not worth exploring, so that the "true" simulator is only run (in the second stage of the kernel) where there is a reasonable chance of accepting proposed values of $\theta$. We show that this approach can be used quite automatically, with the only tuning parameter choice additional to ABC-SMC being the number of particles we wish to carry through to the second stage of the kernel. Applications to stochastic differential equation models and latent doubly intractable distributions are presented.

In this work we propose to fit a sparse logistic regression model by a weakly convex regularized nonconvex optimization problem. The idea is based on the finding that a weakly convex function as an approximation of the $\ell_0$ pseudo norm is able to better induce sparsity than the commonly used $\ell_1$ norm. For a class of weakly convex sparsity inducing functions, we prove the nonconvexity of the corresponding sparse logistic regression problem, and study its local optimality conditions and the choice of the regularization parameter to exclude trivial solutions. Despite the nonconvexity, a method based on proximal gradient descent is used to solve the general weakly convex sparse logistic regression, and its convergence behavior is studied theoretically. Then the general framework is applied to a specific weakly convex function, and a necessary and sufficient local optimality condition is provided. The solution method is instantiated in this case as an iterative firm-shrinkage algorithm, and its effectiveness is demonstrated in numerical experiments by both randomly generated and real datasets.

With the development of neural networks based machine learning and their usage in mission critical applications, voices are rising against the \textit{black box} aspect of neural networks as it becomes crucial to understand their limits and capabilities. With the rise of neuromorphic hardware, it is even more critical to understand how a neural network, as a distributed system, tolerates the failures of its computing nodes, neurons, and its communication channels, synapses. Experimentally assessing the robustness of neural networks involves the quixotic venture of testing all the possible failures, on all the possible inputs, which ultimately hits a combinatorial explosion for the first, and the impossibility to gather all the possible inputs for the second. In this paper, we prove an upper bound on the expected error of the output when a subset of neurons crashes. This bound involves dependencies on the network parameters that can be seen as being too pessimistic in the average case. It involves a polynomial dependency on the Lipschitz coefficient of the neurons activation function, and an exponential dependency on the depth of the layer where a failure occurs. We back up our theoretical results with experiments illustrating the extent to which our prediction matches the dependencies between the network parameters and robustness. Our results show that the robustness of neural networks to the average crash can be estimated without the need to neither test the network on all failure configurations, nor access the training set used to train the network, both of which are practically impossible requirements.

This paper describes a new parameter-free online learning algorithm for changing environments. In comparing against algorithms with the same time complexity as ours, we obtain a strongly adaptive regret bound that is a factor of at least $\sqrt{\log(T)}$ better, where $T$ is the time horizon. Empirical results show that our algorithm outperforms state-of-the-art methods in learning with expert advice and metric learning scenarios.

Recently software development companies started to embrace Machine Learning (ML) techniques for introducing a series of advanced functionality in their products such as personalisation of the user experience, improved search, content recommendation and automation. The technical challenges for tackling these problems are heavily researched in literature. A less studied area is a pragmatic approach to the role of humans in a complex modern industrial environment where ML based systems are developed. Key stakeholders affect the system from inception and up to operation and maintenance. Product managers want to embed "smart" experiences for their users and drive the decisions on what should be built next; software engineers are challenged to build or utilise ML software tools that require skills that are well outside of their comfort zone; legal and risk departments may influence design choices and data access; operations teams are requested to maintain ML systems which are non-stationary in their nature and change behaviour over time; and finally ML practitioners should communicate with all these stakeholders to successfully build a reliable system. This paper discusses some of the challenges we faced in Atlassian as we started investing more in the ML space.

Identifying 3 moss species by deep learning, using the "chopped picture" method Graduate School of Agriculture, Kyoto University, Japan * corresponding author: ise@kais.kyoto-u.ac.jp Abstract In general, object identification tends not to work well on ambiguous, amorphous objects such as vegetation. In this study, we developed a simple but effective approach to identify ambiguous objects and applied the method to several moss species. As a result, the model correctly classified test images with accuracy more than 90%. Using this approach will help progress in computer vision studies. Introduction Especially in recent years, deep learning has become a very effective tool for object identification (Krizhevsky et al. 2012, Szegedy et al. 2015).