Accelerated Diffusion-Based Sampling by the Non-Reversible Dynamics with Skew-Symmetric Matrices.

Langevin dynamics (LD) has been extensively studied theoretically and practically as a basic sampling technique. Recently, the incorporation of non-reversible dynamics into LD is attracting attention because it accelerates the mixing speed of LD. Popular choices for non-reversible dynamics include underdamped Langevin dynamics (ULD), which uses second-order dynamics and perturbations with skew-symmetric matrices. Although ULD has been widely used in practice, the application of skew acceleration is limited although it is expected to show superior performance theoretically. Current work lacks a theoretical understanding of issues that are important to practitioners, including the selection criteria for skew-symmetric matrices, quantitative evaluations of acceleration, and the large memory cost of storing skew matrices. In this study, we theoretically and numerically clarify these problems by analyzing acceleration focusing on how the skew-symmetric matrix perturbs the Hessian matrix of potential functions. We also present a practical algorithm that accelerates the standard LD and ULD, which uses novel memory-efficient skew-symmetric matrices under parallel-chain Monte Carlo settings.

Health checkup and telemedical intervention program for preventive medicine in developing countries: verification study.

BACKGROUND: The prevalence of non-communicable diseases is increasing throughout the world, including developing countries. OBJECTIVE: The intent was to conduct a study of a preventive medical service in a developing country, combining eHealth checkups and teleconsultation as well as assess stratification rules and the short-term effects of intervention. METHODS: We developed an eHealth system that comprises a set of sensor devices in an attaché case, a data transmission system linked to a mobile network, and a data management application. We provided eHealth checkups for the populations of five villages and the employees of five factories/offices in Bangladesh. Individual health condition was automatically categorized into four grades based on international diagnostic standards: green (healthy), yellow (caution), orange (affected), and red (emergent). We provided teleconsultation for orange- and red-grade subjects and we provided teleprescription for these subjects as required. RESULTS: The first checkup was provided to 16,741 subjects. After one year, 2361 subjects participated in the second checkup and the systolic blood pressure of these subjects was significantly decreased from an average of 121 mmHg to an average of 116 mmHg (P<.001). Based on these results, we propose a cost-effective method using a machine learning technique (random forest method) using the medical interview, subject profiles, and checkup results as predictor to avoid costly measurements of blood sugar, to ensure sustainability of the program in developing countries. CONCLUSIONS: The results of this study demonstrate the benefits of an eHealth checkup and teleconsultation program as an effective health care system in developing countries.

A novel deep learning approach for typhoon-induced storm surge modeling through efficient emulation of wind and pressure fields.

Modeling typhoon-induced storm surges requires 10-m wind and sea level pressure fields as forcings, commonly obtained using parametric models or a fully dynamical simulation by numerical weather prediction (NWP) models. The parametric models are generally less accurate than the full-physics models of the NWP, but they are often preferred owing to their computational efficiency facilitating rapid uncertainty quantification. Here, we propose using a deep learning method based on generative adversarial networks (GAN) to translate the parametric model outputs into a more realistic atmospheric forcings structure resembling the NWP model results. Additionally, we introduce lead-lag parameters to incorporate a forecasting feature in our model. Thirty-four historical typhoon events from 1981 to 2012 are selected to train the GAN, followed by storm surge simulations for the four most recent events. The proposed method efficiently transforms the parametric model into realistic forcing fields by a standard desktop computer within a few seconds. The results show that the storm surge model accuracy with forcings generated by GAN is comparable to that of the NWP model and outperforms the parametric model. Our novel GAN model offers an alternative for rapid storm forecasting and can potentially combine varied data, such as those from satellite images, to improve the forecasts further.

Physics-informed deep learning approach for modeling crustal deformation.

The movement and deformation of the Earth's crust and upper mantle provide critical insights into the evolution of earthquake processes and future earthquake potentials. Crustal deformation can be modeled by dislocation models that represent earthquake faults in the crust as defects in a continuum medium. In this study, we propose a physics-informed deep learning approach to model crustal deformation due to earthquakes. Neural networks can represent continuous displacement fields in arbitrary geometrical structures and mechanical properties of rocks by incorporating governing equations and boundary conditions into a loss function. The polar coordinate system is introduced to accurately model the displacement discontinuity on a fault as a boundary condition. We illustrate the validity and usefulness of this approach through example problems with strike-slip faults. This approach has a potential advantage over conventional approaches in that it could be straightforwardly extended to high dimensional, anelastic, nonlinear, and inverse problems.

Machine learning-based tsunami inundation prediction derived from offshore observations.

The world's largest and densest tsunami observing system gives us the leverage to develop a method for a real-time tsunami inundation prediction based on machine learning. Our method utilizes 150 offshore stations encompassing the Japan Trench to simultaneously predict tsunami inundation at seven coastal cities stretching ~100 km along the southern Sanriku coast. We trained the model using 3093 hypothetical tsunami scenarios from the megathrust (Mw 8.0-9.1) and nearby outer-rise (Mw 7.0-8.7) earthquakes. Then, the model was tested against 480 unseen scenarios and three near-field historical tsunami events. The proposed machine learning-based model can achieve comparable accuracy to the physics-based model with ~99% computational cost reduction, thus facilitates a rapid prediction and an efficient uncertainty quantification. Additionally, the direct use of offshore observations can increase the forecast lead time and eliminate the uncertainties typically associated with a tsunami source estimate required by the conventional modeling approach.

Automated acquisition of explainable knowledge from unannotated histopathology images.

Deep learning algorithms have been successfully used in medical image classification. In the next stage, the technology of acquiring explainable knowledge from medical images is highly desired. Here we show that deep learning algorithm enables automated acquisition of explainable features from diagnostic annotation-free histopathology images. We compare the prediction accuracy of prostate cancer recurrence using our algorithm-generated features with that of diagnosis by expert pathologists using established criteria on 13,188 whole-mount pathology images consisting of over 86 billion image patches. Our method not only reveals findings established by humans but also features that have not been recognized, showing higher accuracy than human in prognostic prediction. Combining both our algorithm-generated features and human-established criteria predicts the recurrence more accurately than using either method alone. We confirm robustness of our method using external validation datasets including 2276 pathology images. This study opens up fields of machine learning analysis for discovering uncharted knowledge.

Semisupervised learning for a hybrid generative/discriminative classifier based on the maximum entropy principle.

This paper presents a method for designing semi-supervised classifiers trained on labeled and unlabeled samples. We focus on probabilistic semi-supervised classifier design for multi-class and single-labeled classification problems, and propose a hybrid approach that takes advantage of generative and discriminative approaches. In our approach, we first consider a generative model trained by using labeled samples and introduce a bias correction model, where these models belong to the same model family, but have different parameters. Then, we construct a hybrid classifier by combining these models based on the maximum entropy principle. To enable us to apply our hybrid approach to text classification problems, we employed naive Bayes models as the generative and bias correction models. Our experimental results for four text data sets confirmed that the generalization ability of our hybrid classifier was much improved by using a large number of unlabeled samples for training when there were too few labeled samples to obtain good performance. We also confirmed that our hybrid approach significantly outperformed generative and discriminative approaches when the performance of the generative and discriminative approaches was comparable. Moreover, we examined the performance of our hybrid classifier when the labeled and unlabeled data distributions were different.

A Ranking Approach to Genomic Selection.

BACKGROUND: Genomic selection (GS) is a recent selective breeding method which uses predictive models based on whole-genome molecular markers. Until now, existing studies formulated GS as the problem of modeling an individual's breeding value for a particular trait of interest, i.e., as a regression problem. To assess predictive accuracy of the model, the Pearson correlation between observed and predicted trait values was used. CONTRIBUTIONS: In this paper, we propose to formulate GS as the problem of ranking individuals according to their breeding value. Our proposed framework allows us to employ machine learning methods for ranking which had previously not been considered in the GS literature. To assess ranking accuracy of a model, we introduce a new measure originating from the information retrieval literature called normalized discounted cumulative gain (NDCG). NDCG rewards more strongly models which assign a high rank to individuals with high breeding value. Therefore, NDCG reflects a prerequisite objective in selective breeding: accurate selection of individuals with high breeding value. RESULTS: We conducted a comparison of 10 existing regression methods and 3 new ranking methods on 6 datasets, consisting of 4 plant species and 25 traits. Our experimental results suggest that tree-based ensemble methods including McRank, Random Forests and Gradient Boosting Regression Trees achieve excellent ranking accuracy. RKHS regression and RankSVM also achieve good accuracy when used with an RBF kernel. Traditional regression methods such as Bayesian lasso, wBSR and BayesC were found less suitable for ranking. Pearson correlation was found to correlate poorly with NDCG. Our study suggests two important messages. First, ranking methods are a promising research direction in GS. Second, NDCG can be a useful evaluation measure for GS.

Bayesian model search for mixture models based on optimizing variational bounds.

When learning a mixture model, we suffer from the local optima and model structure determination problems. In this paper, we present a method for simultaneously solving these problems based on the variational Bayesian (VB) framework. First, in the VB framework, we derive an objective function that can simultaneously optimize both model parameter distributions and model structure. Next, focusing on mixture models, we present a deterministic algorithm to approximately optimize the objective function by using the idea of the split and merge operations which we previously proposed within the maximum likelihood framework. Then, we apply the method to mixture of expers (MoE) models to experimentally show that the proposed method can find the optimal number of experts of a MoE while avoiding local maxima.

Modeling of growing networks with directional attachment and communities.

In this paper, we propose a new network growth model and its learning algorithm to more precisely model such a real-world growing network as the Web. Unlike the conventional models, we have incorporated directional attachment and community structure for this purpose. We show that the proposed model exhibits a degree distribution with a power-law tail, which is an important characteristic of many large-scale real-world networks including the Web. Using real Web data, we experimentally show that predictive ability can be improved by incorporating directional attachment and community structure. Also, using synthetic data, we experimentally show that predictive ability can definitely be improved by incorporating community structure.

Visualization of Documents and Concepts in Neuroinformatics with the 3D-SE Viewer.

A new interactive visualization tool is proposed for mining text data from various fields of neuroscience. Applications to several text datasets are presented to demonstrate the capability of the proposed interactive tool to visualize complex relationships between pairs of lexical entities (with some semantic contents) such as terms, keywords, posters, or papers' abstracts. Implemented as a Java applet, this tool is based on the spherical embedding (SE) algorithm, which was designed for the visualization of bipartite graphs. Items such as words and documents are linked on the basis of occurrence relationships, which can be represented in a bipartite graph. These items are visualized by embedding the vertices of the bipartite graph on spheres in a three-dimensional (3-D) space. The main advantage of the proposed visualization tool is that 3-D layouts can convey more information than planar or linear displays of items or graphs. Different kinds of information extracted from texts, such as keywords, indexing terms, or topics are visualized, allowing interactive browsing of various fields of research featured by keywords, topics, or research teams. A typical use of the 3D-SE viewer is quick browsing of topics displayed on a sphere, then selecting one or several item(s) displays links to related terms on another sphere representing, e.g., documents or abstracts, and provides direct online access to the document source in a database, such as the Visiome Platform or the SfN Annual Meeting. Developed as a Java applet, it operates as a tool on top of existing resources.

Parametric embedding for class visualization.

We propose a new method, parametric embedding (PE), that embeds objects with the class structure into a low-dimensional visualization space. PE takes as input a set of class conditional probabilities for given data points and tries to preserve the structure in an embedding space by minimizing a sum of Kullback-Leibler divergences, under the assumption that samples are generated by a gaussian mixture with equal covariances in the embedding space. PE has many potential uses depending on the source of the input data, providing insight into the classifier's behavior in supervised, semisupervised, and unsupervised settings. The PE algorithm has a computational advantage over conventional embedding methods based on pairwise object relations since its complexity scales with the product of the number of objects and the number of classes. We demonstrate PE by visualizing supervised categorization of Web pages, semisupervised categorization of digits, and the relations of words and latent topics found by an unsupervised algorithm, latent Dirichlet allocation.