The purpose of this paper is to provide a comprehensive presentation and interpretation of the Ensemble Kalman Filter (EnKF) and its numerical implementation. The EnKF has a large user group, and numerous publications have discussed applications and theoretical aspects of it. This paper reviews the important results from these studies and also presents new ideas and alternative interpretations which further explain the success of the EnKF. In addition to providing the theoretical framework needed for using the EnKF, there is also a focus on the algorithmic formulation and optimal numerical implementation. A program listing is given for some of the key subroutines. The paper also touches upon specific issues such as the use of nonlinear measurements, in situ profiles of temperature and salinity, and data which are available with high frequency in time. An ensemble based optimal interpolation (EnOI) scheme is presented as a cost-effective approach which may serve as an alternative to the EnKF in some applications. A fairly extensive discussion is devoted to the use of time correlated model errors and the estimation of model bias.

The Ensemble Kalman Filter: Theoretical formulation and practical implementation

Sphere Packings, Lattices and Groups

You must turn in your code as well as output files. Please generate a report that contains the code and ouput in a single readable format. Getting Started  You may want to download Irfanview image viewing software. It handles pretty much any image type, lets you convert, and provides batch processing.  Download the sample images from the class website. The following question operates on the city.jpg image. (a) Perform image smoothing using a 7×7 averaging filter and a Gaussian filter with σ = 0.5 and 3. Compare the outputs. (b) Perform edge enhancement using the Sobel operator (Matlab's default parameters). Repeat using the Laplacian and Laplacian of Guassian operators. Compare the outputs 2. Frequency Domain Filtering The following question operates on the city.jpg image. (a) Find the Fourier transform of the image. Be sure to center the frequencies. (b) Perform image smoothing in the frequency domain using the filters defined in the previous problem. Compare the output images from the two methods (spatial and frequency) and the time for operation. (c) Perform edge enhancement using the filters defined in the previous problem. (d) Define a lowpass filter in the frequency domain with radius of 1/4 the height. Show the result. Repeat with a similar sized Guassian and compare the results. Give the σ parameter you used and show the output transform image. (e) Repeat with a rectangular filter with the same dimension as the ideal lowpass. Compare the results between the ideal filter and the rectangular approximation. 3. Canny Edge Detection (a) Give the convolution kernels for determining the gradient. You may examine the function gradient.m to help with the explanation. (It may be easiest to apply the gradient to an impulse and inspect the results. (b) Implement the simplified version of the Canny edge detector (single scale). The syntax of the function should be where E contains the detected edges, M the smoothed gradient magnitude, A contains the gradient angle, I is the input image, sig is the σ parameter for the smoothing filter, and tau= [τ h , τ l ] is the two element vector containing the hysteresis thresholds. See Algorithm 6.4 for non-maximal suppression and Algorithm 6.5 for hysteresis thresh-olding. (It may be more efficient to implement the hysteresis as edge tracking)

http://ieeexplore.ieee.org/iel5/5/31355/01457626.pdf

Multidimensional digital signal processing

A Genome-Wide Association Study Identifies IL23R as an Inflammatory Bowel Disease Gene.

. We describe the physical model, numerical algorithms, and software structure of a model consisting of the Weather Research and Forecasting (WRF) model, coupled with the fire-spread model (SFIRE) module. In every time step, the fire model inputs the surface wind, which drives the fire, and outputs the heat flux from the fire into the atmosphere, which in turn influences the atmosphere. SFIRE is implemented by the level set method, which allows a submesh representation of the burning region and a flexible implementation of various kinds of ignition. The coupled model is capable of running on a cluster faster than real time even with fine resolution in dekameters. It is available as a part of the Open Wildland Fire Modeling (OpenWFM) environment at http://openwfm.org , which contains also utilities for visualization, diagnostics, and data processing, including an extended version of the WRF Preprocessing System (WPS). The SFIRE code with a subset of the features is distributed with WRF 3.3 as WRF-Fire.

/pdf/coupled-atmosphere-wildland-fire-modeling-with-wrf-3-3-and-2t7zpt6op0.pdf

Coupled atmosphere-wildland fire modeling with WRF 3.3 and SFIRE 2011

Assimilating real-time sensor data into large-scale spatial-temporal simulations, such as simulations of wildfires, is a promising technique for improving simulation results. This asks for advanced data assimilation methods that can work with the complex structures and nonlinear behaviors associated with the simulation models. This article presents a data assimilation framework using Sequential Monte Carlo (SMC) methods for wildfire spread simulations. The models and algorithms of the framework are described, and experimental results are provided. This work demonstrates the feasibility of applying SMC methods to data assimilation of wildfire spread simulations. The developed framework can potentially be generalized to other application areas where sophisticated simulation models are used.

Data assimilation using sequential monte carlo methods in wildfire spread simulation

To defend against an increasing number of sophisticated malware attacks, deep-learning based Malware Detection Systems (MDSs) have become a vital component of our economic and national security. Traditionally, researchers build the single deep learning model using the entire dataset. However, the single deep learning model may not handle the increasingly complex malware data distributions effectively since different sample subspaces representing a group of similar malware may have unique data distribution. In order to further improve the performance of deep learning based MDSs, we propose a Multi-Level Deep Learning System (MLDLS) that organizes multiple deep learning models using the tree structure. Each model in the tree structure of MLDLS was not built on the whole dataset. Instead, each deep learning model focuses on learning a specific data distribution for a particular group of malware and all deep learning models in the tree work together to make a final decision. Consequently, the learning effectiveness of each deep learning model built for one cluster can be improved. Experimental results show that our proposed system performs better than the traditional approach.

A multi-level deep learning system for malware detection

American police departments are beginning to implement social media as a strategy to engage the surrounding communities through various methods, including Facebook and Twitter. The purpose of this paper is to examine the varieties in the use of Facebook and Twitter by local police departments.,This study collected all data between October 1, 2013 and March 31, 2014 from Facebook and Twitter accounts of seven city police departments in the Hampton Roads area of Virginia. These agencies resemble many police departments in the USA, and in total serve a diverse population of approximately 1,435,000. Content analysis and statistical tests are conducted.,Results show that specific types of posts are more engaging for the community. Facebook and Twitter interactions vary depending upon the type of posts, demonstrating that citizens are using Facebook and Twitter to interact in different ways.,The findings presented here give police agencies’ insight on how to appropriately adjust their use of social media to fulfill the needs of the citizens and optimize interactions with the community.,This is the first study to systematically examine and analyze the varieties in the use of social media by traditional American local police departments and their interactions with citizens.

Working with communities on social media: Varieties in the use of Facebook and Twitter by local police

Particle filters are important techniques to support data assimilation for large-scale spatial temporal simulation systems Distributed particle filters improve the performance of particle filtering by distributing particles to multiple processing units (PUs) While different resampling algorithms have been developed for distributed particle filters, less research has been conducted to investigate how to route particles among the PUs after resampling in effective and efficient manners This paper develops particle routing policies in distributed particle filters with both the centralized resampling and the distributed resampling The developed routing policies are evaluated from the aspects of the communication cost and the data assimilation accuracy based on an application of data assimilation for large-scale wildfire spread simulations

Particle Routing in Distributed Particle Filters for Large-Scale Spatial Temporal Systems

Wildfire propagation is a complex process influenced by many factors. Simulation models of wildfire spread, such as DEVS-FIRE, are important tools for studying fire behavior. This paper presents how the sequential Monte Carlo methods, i.e., particle filters, can work together with DEVS-FIRE for better simulation and prediction of wildfire. We define an application framework of particle filters for the problem of wildfire spread using the DEVS-FIRE model, and discuss several applications. A case study example is provided and preliminary results are presented.

/pdf/towards-applications-of-particle-filters-in-wildfire-spread-45iltv78u4.pdf

Feng Gu

Papers

Data assimilation using sequential monte carlo methods in wildfire spread simulation

A multi-level deep learning system for malware detection

Working with communities on social media: Varieties in the use of Facebook and Twitter by local police

Particle Routing in Distributed Particle Filters for Large-Scale Spatial Temporal Systems

Towards applications of particle filters in wildfire spread simulation