We develop a personalized real time risk scoring algorithm that provides timely and granular assessments for the clinical acuity of ward patients based on their (temporal) lab tests and vital signs. Heterogeneity of the patients population is captured via a hierarchical latent class model. The proposed algorithm aims to discover the number of latent classes in the patients population, and train a mixture of Gaussian Process (GP) experts, where each expert models the physiological data streams associated with a specific class. Self-taught transfer learning is used to transfer the knowledge of latent classes learned from the domain of clinically stable patients to the domain of clinically deteriorating patients. For new patients, the posterior beliefs of all GP experts about the patient's clinical status given her physiological data stream are computed, and a personalized risk score is evaluated as a weighted average of those beliefs, where the weights are learned from the patient's hospital admission information. Experiments on a heterogeneous cohort of 6,313 patients admitted to Ronald Regan UCLA medical center show that our risk score outperforms the currently deployed risk scores, such as MEWS and Rothman scores.

We propose an efficient method for approximating natural gradient descent in neural networks which we call Kronecker-Factored Approximate Curvature (K-FAC). K-FAC is based on an efficiently invertible approximation of a neural network's Fisher information matrix which is neither diagonal nor low-rank, and in some cases is completely non-sparse. It is derived by approximating various large blocks of the Fisher (corresponding to entire layers) as being the Kronecker product of two much smaller matrices. While only several times more expensive to compute than the plain stochastic gradient, the updates produced by K-FAC make much more progress optimizing the objective, which results in an algorithm that can be much faster than stochastic gradient descent with momentum in practice. And unlike some previously proposed approximate natural-gradient/Newton methods which use high-quality non-diagonal curvature matrices (such as Hessian-free optimization), K-FAC works very well in highly stochastic optimization regimes. This is because the cost of storing and inverting K-FAC's approximation to the curvature matrix does not depend on the amount of data used to estimate it, which is a feature typically associated only with diagonal or low-rank approximations to the curvature matrix.

We develop a general approach to valid inference after model selection. At the core of our framework is a result that characterizes the distribution of a post-selection estimator conditioned on the selection event. We specialize the approach to model selection by the lasso to form valid confidence intervals for the selected coefficients and test whether all relevant variables have been included in the model.

Network analysis offers a new set of techniques to tackle the persistent and growing problem of complex fraud. Network analysis supplements traditional techniques by providing a mechanism to bridge investigative and analytics methods. Beyond base visualization, network analysis provides a standardized platform for complex fraud pattern storage and retrieval, pattern discovery and detection, statistical analysis, and risk scoring. This article gives an overview of the main challenges and demonstrates a promising approach using a hands-on example. With swelling globalization, advanced digital communication technology, and international financial deregulation, fraud investigators face a daunting battle against increasingly sophisticated fraudsters.

Warren Buffett isn't known for blown calls. But the billionaire investment guru and head of the holding company Berkshire Hathaway has become increasingly touchy around one of his top positions: International Business Machines. "We feel fine or we won't own it. We've never sold a share of IBM," Buffett told CNBC in an interview Monday. The fact that shares in the computing giant have fallen some 15 percent since he first started building his stake in the company in early 2011 hasn't phased him.

This is actually a good question. For a categorical variable, can the model say that some veles are significant, some levels are not. Typically after a regression we look at the ANOVA (Analysis of Variance) table. There we have 1 row per independent variable. In other words, in My example we will see a single row corresponding to the variable COLOR (as opposed to say 2 rows for I_green and I_blue).

Deep neural networks (DNNs) are capable of learning to identify shapes, so "we're on the right track in developing machines with a visual system and vocabulary as flexible and versatile as ours," say KU Leuven researchers. "For the first time, a dramatic increase in performance has been observed on object and scene categorization tasks, quickly reaching performance levels rivaling humans," they note in an open-access paper in PLOS Computational Biology. Categorization accuracy for models created by three DNNs (CaffeNet, VGG-19, and GoggLeNet) for three types of images (color, grayscaled, silhouette). For each type, mean human performance is indicated by a gray horizontal line, with the gray surrounding band depicting 95% confidence intervals. Error bars (vertical black lines) depict 95% confidence intervals.

Over the last two decades, California and the federal government have faced harsh criticism for failing to take stronger actions to stop a highly contagious disease that has killed millions of trees along coastal regions from Big Sur to portions of Oregon. Now, a new computer modeling study suggests that the "sudden oak death" epidemic, which emerged in 1995, has grown too big and is spreading too fast to eradicate statewide. The analysis is the first to integrate knowledge of the pathogen with topography, weather and resources like government budgets to predict the likely effects of various management strategies over such a large area -- in this case, California's 163,707 square miles of land. The results are somewhat hopeful: Because the epidemic's growth rate increases with its size, focusing on restoring and treating small, local forests is now the most practical and cost-effective option for managing the destructive fungus, Phtophthora ramorum. The findings were published Monday in the Proceedings of the National Academy of Sciences.

ANALYSIS: Ray Dalio, the founder of Bridgewater Associates, the largest hedge fund in the world, hired IBM's Watson programming team to incorporate artificial intelligence into the investment process. Later this year, I will be doing something similar, so stay tuned to learn more about this exciting endeavour. Today we have Jamie Wise from BUZZ Indexes based in Toronto who recently launched an Artificial Intelligence driven ETF based index fund BUZ (US). The investment process uses natural language processing software and artificial intelligence to scour the internet for tweets, blogs, and searches for company names. One of the most interesting things is that the software is able distinguish between someone typing Amazon when talking about a bad product they bought on line and AMZN and a bad outlook for the stock.

In "The Three Rs and an S," a recent op-ed in The Baltimore Sun, you and Norman Augustine write that the Next Generation Science Standards will teach students how to take on a scientific manner of thinking. Why is there a greater emphasis now on this approach to learning than in years past? The future jobs for this millennial generation will not look like the jobs of the last 40 or 50 years. They're not going to be large segments of people working in factories. Really, what's happened is that the business community itself has sort of shifted where it sees efficiency and productivity occurring and, because of this shift, the jobs are going to shift.