The majority of medical documents and electronic health records (EHRs) are in text format that poses a challenge for data processing and finding relevant documents. Looking for ways to automatically retrieve the enormous amount of health and medical knowledge has always been an intriguing topic. Powerful methods have been developed in recent years to make the text processing automatic. One of the popular approaches to retrieve information based on discovering the themes in health & medical corpora is topic modeling, however, this approach still needs new perspectives. In this research we describe fuzzy latent semantic analysis (FLSA), a novel approach in topic modeling using fuzzy perspective. FLSA can handle health & medical corpora redundancy issue and provides a new method to estimate the number of topics. The quantitative evaluations show that FLSA produces superior performance and features to latent Dirichlet allocation (LDA), the most popular topic model.

DUBLIN, IRELAND - MAY 11: (Photo by Paul Faith - WPA Pool/Getty Images) 12. Active Scheduling A.I. helpers like Siri and Google's new Assistant have always at least attempted to help their users schedule their various appointments, but it's only with the very recent introduction of artificial intelligence that these projects have managed to do more than remember past appointments and repeat them. Now, and especially in the near future, A.I. can read and understand your conversations to actively pull out scheduling info.

In the last ten years, artificial intelligence has changed the world in subtle but sweeping ways, but it's got nothing on the coming decade, if you look at what's being developed today. Voice recognition on every smartphone were simple proofs of concept. Over the next 10 years, artificial intelligence will make more progress than in the fifty before it, combined. With countless quickly oncoming applications to business, government, and personal life, its influence will soon touch absolutely every aspect of our lives. Here are 25 surprising ways life and society that will be forever changed by artificial intelligence over the coming decade.

We propose a flexible convex relaxation for the phase retrieval problem that operates in the natural domain of the signal. Therefore, we avoid the prohibitive computational cost associated with "lifting" and semidefinite programming (SDP) in methods such as PhaseLift and compete with recently developed non-convex techniques for phase retrieval. We relax the quadratic equations for phaseless measurements to inequality constraints each of which representing a symmetric "slab". Through a simple convex program, our proposed estimator finds an extreme point of the intersection of these slabs that is best aligned with a given anchor vector. We characterize geometric conditions that certify success of the proposed estimator. Furthermore, using classic results in statistical learning theory, we show that for random measurements the geometric certificates hold with high probability at an optimal sample complexity. Phase transition of our estimator is evaluated through simulations. Our numerical experiments also suggest that the proposed method can solve phase retrieval problems with coded diffraction measurements as well.

We propose Edward, a Turing-complete probabilistic programming language. Edward defines two compositional representations---random variables and inference. By treating inference as a first class citizen, on a par with modeling, we show that probabilistic programming can be as flexible and computationally efficient as traditional deep learning. For flexibility, Edward makes it easy to fit the same model using a variety of composable inference methods, ranging from point estimation to variational inference to MCMC. In addition, Edward can reuse the modeling representation as part of inference, facilitating the design of rich variational models and generative adversarial networks. For efficiency, Edward is integrated into TensorFlow, providing significant speedups over existing probabilistic systems. For example, we show on a benchmark logistic regression task that Edward is at least 35x faster than Stan and 6x faster than PyMC3. Further, Edward incurs no runtime overhead: it is as fast as handwritten TensorFlow.

Probabilistic modeling is a powerful approach for analyzing empirical information. We describe Edward, a library for probabilistic modeling. Edward's design reflects an iterative process pioneered by George Box: build a model of a phenomenon, make inferences about the model given data, and criticize the model's fit to the data. Edward supports a broad class of probabilistic models, efficient algorithms for inference, and many techniques for model criticism. The library builds on top of TensorFlow to support distributed training and hardware such as GPUs. Edward enables the development of complex probabilistic models and their algorithms at a massive scale.

When approximating a black-box function, sampling with active learning focussing on regions with non-linear responses tends to improve accuracy. We present the FLOLA-Voronoi method introduced previously for deterministic responses, and theoretically derive the impact of output uncertainty. The algorithm automatically puts more emphasis on exploration to provide more information to the models.

If no CONTROL slot is associated with a prototype, the Interpreter will attempt to fill in values for the prototype components in the order of their Importance measures. When all of the clauses in the CONTROL slot have been executed and the prototype has been instantiated, a decision is madel as to whether the prototype should be confirmed as matching the data in the case. The system then checks either the IF-CONFIRMED slot or the IF-DISPROVED slot to determine what should be done next. Similarly, the ACTION slot specifies stops to be taken for a confirmed prototype during the clean-up stage.

We describe the development and (partial) Implementation of an "automated consultant" to advise non-expert engineers In the use of a general-purpose structural analysis program. The analysis program numerically simulates the behavior of a physical structure subjected to various mechanical loading conditions. The automated consultant, called SACON (Structural Analysis CONsultant), Is based on a version of the MYCIN program [Shortliffe74], originally developed to advise physicians In the diagnosis and treatment of infectious diseases. The domain-specific knowledge in MYCIN Is represented as situation-action rules, and is kept independent of the "inference engine" that uses the rules. By substituting structural engineering knowledge for the medical knowledge, the program was converted easily from the domain of Infectious diseases to the domain of structural analysis.

An important subclass of hybrid Bayesian networks are those that represent Conditional Linear Gaussian (CLG) distributions --- a distribution with a multivariate Gaussian component for each instantiation of the discrete variables. In this paper we explore the problem of inference in CLGs. We show that inference in CLGs can be significantly harder than inference in Bayes Nets. In particular, we prove that even if the CLG is restricted to an extremely simple structure of a polytree in which every continuous node has at most one discrete ancestor, the inference task is NP-hard.To deal with the often prohibitive computational cost of the exact inference algorithm for CLGs, we explore several approximate inference algorithms. These algorithms try to find a small subset of Gaussians which are a good approximation to the full mixture distribution. We consider two Monte Carlo approaches and a novel approach that enumerates mixture components in order of prior probability. We compare these methods on a variety of problems and show that our novel algorithm is very promising for large, hybrid diagnosis problems.