Browse Source

Added conference poster template

master
Micah Halter 3 months ago
parent
commit
ef3f225203
25 changed files with 775 additions and 30 deletions
  1. +2
    -30
      README.md
  2. +16
    -0
      conference-poster/README.md
  3. BIN
      conference-poster/figs/endemic_graph.png
  4. BIN
      conference-poster/figs/gtri.png
  5. +15
    -0
      conference-poster/figs/logo.dot
  6. +56
    -0
      conference-poster/figs/logo.svg
  7. BIN
      conference-poster/figs/lotka-volterra.png
  8. BIN
      conference-poster/figs/malaria-diffeq.png
  9. BIN
      conference-poster/figs/malaria.png
  10. BIN
      conference-poster/figs/result_graph.png
  11. BIN
      conference-poster/figs/result_graph_zoom.png
  12. BIN
      conference-poster/figs/semanticmodels.png
  13. BIN
      conference-poster/figs/volterra-diffeq.png
  14. BIN
      conference-poster/figs/volterra-ross.png
  15. +501
    -0
      conference-poster/poster.html
  16. BIN
      conference-poster/poster.pdf
  17. +127
    -0
      conference-poster/poster.rmd
  18. +27
    -0
      conference-poster/refs.bib
  19. +31
    -0
      letterhead/README.md
  20. +0
    -0
      letterhead/gtri-ltr.sty
  21. +0
    -0
      letterhead/letter.md
  22. +0
    -0
      letterhead/letter.pdf
  23. +0
    -0
      letterhead/letter.tex
  24. +0
    -0
      letterhead/letter_template.tex
  25. +0
    -0
      letterhead/logo.png

+ 2
- 30
README.md View File

@@ -1,31 +1,3 @@
# GTRI-Letterhead
# GTRI Templates

This is a LaTeX and pandoc markdown template for the Georgia Tech Research Institute Letterhead.

## Usage

LaTeX template:

`latexmk --pdf letter.tex`

pandoc markdown template:

`pandoc --template letter_template.tex letter.md -o letter.pdf`

pandoc template options:

| option | details | required |
|-------------|---------------------------|--------------------------------------------|
| `name` | sender name | required |
| `company` | sender company | required |
| `position` | sender position | required |
| `address` | sender street address | required |
| `citystate` | sender city, state zip | required |
| `phone` | sender phone number | optional |
| `email` | sender email | optional |
| `website` | sender website | optional |
| `recipient` | receiver name and address | required |
| `date` | date of letter | optional, default today |
| `opening` | letter greeting | optional, default "To Whom It May Concern" |
| `closing` | letter closing | optional, default "Sincerely" |
| `options` | `logo` or `no-logo` | optional, default `logo` |
This is a collection of useful GTRI LaTeX, pandoc, and rmarkdown templates.

+ 16
- 0
conference-poster/README.md View File

@@ -0,0 +1,16 @@
# GTRI Conference Poster Template


This is an rmarkdown template for a Georgia Tech Research Institute Conference Poster.

This is based on [posterdown](https://github.com/brentthorne/posterdown), see
their wiki for more configuration options.


## Usage

Compile in R with `rmarkdown::render("poster.rmd")`

Compile from terminal prompt with `echo "rmarkdown::render('poster.rmd')" | R -q --vanilla`

To get PDF, print to file from a browser with page size of "Architectural E"

BIN
conference-poster/figs/endemic_graph.png View File

Before After
Width: 895  |  Height: 597  |  Size: 106KB

BIN
conference-poster/figs/gtri.png View File

Before After
Width: 11118  |  Height: 2409  |  Size: 409KB

+ 15
- 0
conference-poster/figs/logo.dot View File

@@ -0,0 +1,15 @@
digraph G {
graph [bgcolor="transparent", nodesep=".3", ranksep=".3"];
node[shape=circle, style=filled]
//rankdir=LR
code1 [label="", color="royalblue3", fillcolor="royalblue1"]
code2 [label="", color="brown3", fillcolor="indianred"]
m1 [label="", color="forestgreen", fillcolor="mediumseagreen"]
m2 [label="", color="mediumorchid3", fillcolor="orchid3"]
{rank=same code1, m1}
{rank=same code2, m2}
code1 -> m1 [color="white"]
code1 -> code2 [color="white"]
code2 -> m2 [color="white"]
m1 -> m2 [color="white"]
}

+ 56
- 0
conference-poster/figs/logo.svg View File

@@ -0,0 +1,56 @@
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
<!-- Generated by graphviz version 2.43.20190912.0211 (20190912.0211)
-->
<!-- Title: G Pages: 1 -->
<svg width="102pt" height="102pt"
viewBox="0.00 0.00 102.00 102.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 98)">
<title>G</title>
<!-- code1 -->
<g id="node1" class="node">
<title>code1</title>
<ellipse fill="#4876ff" stroke="#3a5fcd" cx="18" cy="-76" rx="18" ry="18"/>
</g>
<!-- code2 -->
<g id="node2" class="node">
<title>code2</title>
<ellipse fill="indianred" stroke="#cd3333" cx="18" cy="-18" rx="18" ry="18"/>
</g>
<!-- code1&#45;&gt;code2 -->
<g id="edge2" class="edge">
<title>code1&#45;&gt;code2</title>
<path fill="none" stroke="white" d="M18,-57.89C18,-54.2 18,-50.21 18,-46.24"/>
<polygon fill="white" stroke="white" points="21.5,-46 18,-36 14.5,-46 21.5,-46"/>
</g>
<!-- m1 -->
<g id="node3" class="node">
<title>m1</title>
<ellipse fill="mediumseagreen" stroke="forestgreen" cx="76" cy="-76" rx="18" ry="18"/>
</g>
<!-- code1&#45;&gt;m1 -->
<g id="edge1" class="edge">
<title>code1&#45;&gt;m1</title>
<path fill="none" stroke="white" d="M36.12,-76C39.99,-76 43.85,-76 47.71,-76"/>
<polygon fill="white" stroke="white" points="47.93,-79.5 57.93,-76 47.93,-72.5 47.93,-79.5"/>
</g>
<!-- m2 -->
<g id="node4" class="node">
<title>m2</title>
<ellipse fill="#cd69c9" stroke="#b452cd" cx="76" cy="-18" rx="18" ry="18"/>
</g>
<!-- code2&#45;&gt;m2 -->
<g id="edge3" class="edge">
<title>code2&#45;&gt;m2</title>
<path fill="none" stroke="white" d="M36.12,-18C39.99,-18 43.85,-18 47.71,-18"/>
<polygon fill="white" stroke="white" points="47.93,-21.5 57.93,-18 47.93,-14.5 47.93,-21.5"/>
</g>
<!-- m1&#45;&gt;m2 -->
<g id="edge4" class="edge">
<title>m1&#45;&gt;m2</title>
<path fill="none" stroke="white" d="M76,-57.89C76,-54.2 76,-50.21 76,-46.24"/>
<polygon fill="white" stroke="white" points="79.5,-46 76,-36 72.5,-46 79.5,-46"/>
</g>
</g>
</svg>

BIN
conference-poster/figs/lotka-volterra.png View File

Before After
Width: 765  |  Height: 766  |  Size: 112KB

BIN
conference-poster/figs/malaria-diffeq.png View File

Before After
Width: 806  |  Height: 494  |  Size: 48KB

BIN
conference-poster/figs/malaria.png View File

Before After
Width: 858  |  Height: 526  |  Size: 136KB

BIN
conference-poster/figs/result_graph.png View File

Before After
Width: 584  |  Height: 392  |  Size: 30KB

BIN
conference-poster/figs/result_graph_zoom.png View File

Before After
Width: 590  |  Height: 393  |  Size: 25KB

BIN
conference-poster/figs/semanticmodels.png View File

Before After
Width: 1519  |  Height: 255  |  Size: 52KB

BIN
conference-poster/figs/volterra-diffeq.png View File

Before After
Width: 620  |  Height: 331  |  Size: 33KB

BIN
conference-poster/figs/volterra-ross.png View File

Before After
Width: 1233  |  Height: 1233  |  Size: 270KB

+ 501
- 0
conference-poster/poster.html View File

@@ -0,0 +1,501 @@
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta charset="utf-8" />
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<meta name="generator" content="pandoc" />
<meta name="viewport" content="width=device-width, initial-scale=1">



<title>poster.utf8</title>






<!--
Font-awesome icons ie github or twitter
-->
<link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.8.1/css/all.css" integrity="sha384-50oBUHEmvpQ+1lW4y57PTFmhCaXp0ML5d60M1M7uH2+nqUivzIebhndOJK28anvf" crossorigin="anonymous">
<link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.8.1/css/brands.css" integrity="sha384-n9+6/aSqa9lBidZMRCQHTHKJscPq6NW4pCQBiMmHdUCvPN8ZOg2zJJTkC7WIezWv" crossorigin="anonymous">

<!--
Google fonts api stuff
-->
<link href='https://fonts.googleapis.com/css?family=Rasa' rel='stylesheet'>
<link href='https://fonts.googleapis.com/css?family=Rasa' rel='stylesheet'>

<!--
Here are the required style attributes for css to make this poster work :)
-->
<style>
@page {
size: 36in 48in;
margin: 0;
padding: 0;
}
body {
margin: 0px;
padding: 0px;
width: 36in;
height: 48in;
text-align: justify;
font-size: 45px;
line-height: 1.05;
}
/* RMarkdown Class Styles */
/* center align leaflet map,
from https://stackoverflow.com/questions/52112119/center-leaflet-in-a-rmarkdown-document */
.html-widget {
margin: auto;
position: sticky;
margin-top: 2cm;
margin-bottom: 2cm;
}
.leaflet.html-widget.html-widget-static-bound.leaflet-container.leaflet-touch.leaflet-fade-anim.leaflet-grab.leaflet-touch-drag.leaflet-touch-zoom {
position: sticky;
width: 100%;
}
pre.sourceCode.r {
background-color: #dddddd40;
border-radius: 4mm;
padding: 4mm;
width: 75%;
margin: auto;
margin-top: 1em;
margin-bottom: 1em;
/* align-items: center; */
}
code.sourceCode.r{
background-color: transparent;
font-size: 20pt;
border-radius: 2mm;
}
code {
font-size: 25pt;
font-family: monospace;
background-color: #101B4924;
color: #101B49;
padding: 1.2mm;
line-height: 1;
border-radius: 2mm;
}
caption {
margin-bottom: 10px;
font-size: 20pt;
font-style: italic;
}

tbody tr:nth-child(odd) {
background-color: #101B4920;
}
.table>thead>tr>th, .table>tbody>tr>th, .table>tfoot>tr>th, .table>thead>tr>td, .table>tbody>tr>td, .table>tfoot>tr>td{
border-spacing: 0;
font-size: 40%;
border-style: none;
padding-top: 15px;
padding-bottom: 15px;
padding-right: 1em;
padding-left: 1em;
line-height: 1em;
}
table {
margin: auto;
}
th {
padding-left: 5mm;
padding-right: 5mm;
}
.caption {
font-size: 20pt;
font-style: italic;
padding-top: 0;
}
.references {
font-size: 20px;
line-height: 90%;
}
/* Create three unequal columns that floats next to each other */
.column {
float: left;
padding: 0px;
}
.outer {
width: 36in;
height: calc(48in * (1 - 0.2 - 0.1 - 0.01) );
-webkit-column-count: 3; /* Chrome, Safari, Opera */
-moz-column-count: 3; /* Firefox */
column-count: 3;
-webkit-column-fill: auto;
-moz-column-fill: auto;
column-fill: auto;
column-gap: 0;
padding-left: 0cm;
padding-right: 0cm;
/* -webkit-column-rule-width: 50%;
-moz-column-rule-width: 50%;
column-rule-width: 50%; */
-webkit-column-rule-style: none;
-moz-column-rule-style: none;
column-rule-style: none;
-webkit-column-rule-color: black;
-moz-column-rule-color: black;
column-rule-color: black;
background-color: #ffffff;
font-family: Rasa;
margin-top: calc(48in * 0.2 );
padding-top: 1em;
padding-bottom: 1em;
}
span.citation {
color: #101B49;
font-weight: bold;
}
a {
text-decoration: none;
color: #101B49;
}
#title {
font-size: 90pt;
text-align: left;
margin: 0;
line-height: 98%;
border-bottom: 0;
font-weight: normal;
background: 0;
}
#author {
color: #101B49;
margin: 0;
line-height: 85%;
font-size: 1.17em;
}
#affiliation {
padding-top: 0.1em;
color: ;
font-style: italic;
font-size: 25px;
margin: 0;
}
sup {
color: #cc0000;
}
.affiliation sup {
font-size: 20px;
}
.author {
text-align: left;
}
.author sup {
font-size: 30px;
}
.author_extra {
color: #101B49;
margin: 0;
line-height: 85%;
font-size: 35px;
text-align: left;
}
.outer h1, h2, h3, h4, h5, h6 {
text-align: center;
margin: 0;
font-weight: bold;
}
.section h1 {
text-align:center;
padding-bottom:5px;
background:
linear-gradient(
to left,
#ffffff 1%,
#ffffff 20%,
#101B4975 33%,
#101B49 50%,
#101B4975 66%,
#ffffff 80%,
#ffffff 99%
)
left
bottom
#ffffff
no-repeat;
background-size:100% 5px ;
margin-top: 0.5em;
margin-bottom: 0.5em;
}
.outer h2 {
text-align: center;
}
.outer p, .level2 {
color: #000000;
}
.outer ol {
padding-left: 8%;
padding-right: 8%;
text-align: left;
}
.main {
width: 36in;
height: calc(48in * 0.2);
position: absolute;
background-color: #101B49;
color: #ffffff90;
font-family: Rasa;
background-image: linear-gradient(#101B49 50%,#101B49);
}
.main strong {
color: #ffffff;
}
.main strong > sup {
color: #ffffff;
}
.main sup {
color: #ffffff90;
}
#main-img-left {
width: 10%;
left: 0.5in;
bottom: 0.2in;
position: absolute;
}
#main-img-center {
width: 10%;
left: calc(36in * 0.45);
bottom: 0.5in;
position: absolute;
}
#main-img-right {
width: 10%;
right: 0.5in;
bottom: 0.2in;
position: absolute;
}
.main p {
font-size: 150px;
font-family: Rasa;
text-align: center;
margin: 0;
position: absolute;
top: 50%;
-ms-transform: translateY(-50%);
transform: translateY(-50%);
margin-left: 1em;
}
.fab {
color: #00000030;
font-size: 25px;
}
.twitter, i {
color: #00000030;
font-size: 35px;
text-decoration: none;
}
a.email {
text-decoration: none;
color: #00000030;
font-size: 35px;
}
.envelope {
color: #00000030;
font-size: 5px;
text-decoration: none;
}
.poster_wrap {
width: 36in;
height: 48in;
padding: 0cm;
}
.main_bottom {
width: 36in;
height: calc(48in * 0.1);
margin-top: calc(48in * (1 - 0.1));
position: absolute;
background-color: #101B49;
background-image: linear-gradient(#101B49 10%, #101B49);
}
.section {
padding-left: 10mm;
padding-right: 10mm;
}
span > #tab:mytable {
font-weight: bold;
}
.orcid img {
width: 3%;
}
.emphasis {
background-color: #008080;
color: #ffffff;
border: solid #0b2045 3mm;
margin: 1em;
padding-left: 0;
padding-right: 0;
}
.emphasis h1 {
font-weight: bold;
background: none;
background-color: #0b2045;
padding-bottom: 5mm;
padding-top: 1mm;
margin-top: -1mm;
margin-right: -1mm;
margin-left: -1mm;
}
.emphasis blockquote {
border: 0;
}
.emphasis ol {
padding: 0;
padding-left: 8%;
font-size: 100%;
font-weight: bold;
}
.emphasis p {
color: #ffffff;
}
</style>
</head>
<body>


<div class="poster_wrap">

<div class="column outer">
<div class="section">

<h3 id="author" class="author">

<strong>Micah Halter</strong><sup> 1</sup><br>

<a class='envelope'><i class="fas fa-envelope"></i></a> <a href="mailto:Micah.Halter@gtri.gatech.edu" class="email">Micah.Halter@gtri.gatech.edu</a> <br>
</h3>

<h5 id="author_extra", class="author_extra">
Kun Cao<sup>1</sup>
James Fairbanks<sup>1</sup>
</h5>


<p id="affiliation" class="affiliation">
<sup>1</sup> Georgia Tech Research Institute
</p>
</div>

<style>
.main p {
margin-left: 0em;
}
#main-img-left {
width: 35%;
bottom: 1.25in;
}
#main-img-right {
width: 30%;
bottom: 1.25in;
}
</style>
<div id="introduction" class="section level1">
<h1>Introduction</h1>
<p>Scientific progress comes from adapting and extending models from prior work to address new problems. The task of modifying existing source code and performing semantically sound model fusion can be difficult as the models scale, and become more complex. We propose <code>SemanticModels.jl</code>, a category theory-based framework for defining meta-modeling tasks and semantic information extraction. Our objectives are threefold:</p>
<ol style="list-style-type: decimal">
<li>Convert scientific models to a universal wiring diagram representation <span class="citation">(Fong <a href="#ref-fong" role="doc-biblioref">2018</a>)</span>;</li>
<li>Use this representation to augment and compose these models at domain-level semantics;</li>
<li>Generate executable code and visualize results.</li>
</ol>
</div>
<div id="methods" class="section level1">
<h1>Methods</h1>
<p>We leveraged techniques from static and dynamic program analysis to process executable versions of scientific models to extract the mathematical representation of the Lotka-Volterra Predator-Prey <span class="citation">(Volterra <a href="#ref-volterra" role="doc-biblioref">1926</a>)</span> and the MacDonald Ross Disease Model <span class="citation">(Smith <a href="#ref-ross" role="doc-biblioref">2012</a>)</span>. Next we performed meta-modeling tasks defined as domain specific transformations to augment and compose the two models. Finally, we can visualize the new model as well as generate executable code and visualize the simulation’s results.</p>
</div>
<div id="results" class="section level1">
<h1>Results</h1>
<p>Figures <a href="#fig:volterra">2</a> and <a href="#fig:ross">3</a> show the universal, domain specific wiring diagrams that domain scientists can manipulate to create new models. Figure <a href="#fig:ross">3</a> also illustrates the amount of math that is contained in the simple Malaria wiring diagram.
These wiring diagrams allow us to perform model augmentation and fusion at the semantic level, and the fusion of the differential equations is handled automatically.
We then augmented the Lotka-Volterra model from Figure <a href="#fig:volterra">2</a> to model mutual predation and composed it with the Ross Malaria model from Figure <a href="#fig:ross">3</a> to get a model that simulates both, shown in Figure <a href="#fig:volterraross">4</a>.
Finally we generated the code to simulate the new model, and the results are shown in Figures <a href="#fig:results">5</a> and <a href="#fig:resultszoom">6</a>.</p>
<div class="figure" style="text-align: center"><span id="fig:volterradiffeq"></span>
<img src="figs/volterra-diffeq.png" alt="The full set of differential equations of the Lotka-Volterra model that we want to represent as a mathematical structure grounded in semantic knowledge." width="80%" />
<p class="caption">
Figure 1: The full set of differential equations of the Lotka-Volterra model that we want to represent as a mathematical structure grounded in semantic knowledge.
</p>
</div>
<div class="figure" style="text-align: center"><span id="fig:volterra"></span>
<img src="figs/lotka-volterra.png" alt="**Top**: Lotka-Volterra predator-prey model represented as a semantically rich wiring diagram;&lt;br /&gt;**Bottom**: Lotka-Volterra Predator-Prey Model represented as a structured mathematical object" width="95%" />
<p class="caption">
Figure 2: <strong>Top</strong>: Lotka-Volterra predator-prey model represented as a semantically rich wiring diagram;<br /><strong>Bottom</strong>: Lotka-Volterra Predator-Prey Model represented as a structured mathematical object
</p>
</div>
<p><br /></p>
<div class="figure" style="text-align: center"><span id="fig:ross"></span>
<img src="figs/malaria.png" alt="**Left**: Ross Malaria model represented as a wiring diagram;&lt;br /&gt;**Right**: The full set of differential equations represented by the wiring diagram on the left." width="49%" /><img src="figs/malaria-diffeq.png" alt="**Left**: Ross Malaria model represented as a wiring diagram;&lt;br /&gt;**Right**: The full set of differential equations represented by the wiring diagram on the left." width="49%" />
<p class="caption">
Figure 3: <strong>Left</strong>: Ross Malaria model represented as a wiring diagram;<br /><strong>Right</strong>: The full set of differential equations represented by the wiring diagram on the left.
</p>
</div>
<p><br /></p>
<div class="figure" style="text-align: center"><span id="fig:volterraross"></span>
<img src="figs/volterra-ross.png" alt="**Top**: The generated Lotka-Volterra and Malaria disease spread model as a wiring diagram;&lt;br /&gt;**Bottom**: The generated Lotka-Volterra and Malaria disease spread model as a mathematical structure." width="90%" />
<p class="caption">
Figure 4: <strong>Top</strong>: The generated Lotka-Volterra and Malaria disease spread model as a wiring diagram;<br /><strong>Bottom</strong>: The generated Lotka-Volterra and Malaria disease spread model as a mathematical structure.
</p>
</div>
<div class="figure" style="text-align: center"><span id="fig:results"></span>
<img src="figs/result_graph.png" alt="The results of the mutual predation and Malaria model showing the predator-prey interaction between the mosquitoes and birds (purple and yellow lines respectively)" width="95%" />
<p class="caption">
Figure 5: The results of the mutual predation and Malaria model showing the predator-prey interaction between the mosquitoes and birds (purple and yellow lines respectively)
</p>
</div>
<p><br /></p>
<div class="figure" style="text-align: center"><span id="fig:resultszoom"></span>
<img src="figs/result_graph_zoom.png" alt="Zooming into the beginning of the graph in Figure \@ref(fig:results) shows the initial role Malaria had on the system." width="95%" />
<p class="caption">
Figure 6: Zooming into the beginning of the graph in Figure <a href="#fig:results">5</a> shows the initial role Malaria had on the system.
</p>
</div>
<p><br /></p>
</div>
<div id="conclusions" class="section level1">
<h1>Conclusions</h1>
<ol style="list-style-type: decimal">
<li><code>SemanticModels.jl</code> provides a framework for scientists to easily augment and transform existing scientific models.</li>
<li>Using a universal representation allows for models to be easily composed across scientific domains.</li>
<li>New models can automatically generate executable code and produce verifiable results.</li>
<li>Our representation can be adapted to curate more types of models that can then be used in the same workflow.</li>
</ol>
</div>
<div id="acknowledgments" class="section level1">
<h1>Acknowledgments</h1>
<p>This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Agreement No. HR00111990008.</p>
</div>
<div id="references" class="section level1 unnumbered">
<h1>References</h1>
<div id="refs" class="references">
<div id="ref-fong">
<p>Fong, Brendan. 2018. “Seven Sketches in Compositionality: An Invitation to Applied Category Theory.”</p>
</div>
<div id="ref-ross">
<p>Smith, Katherine E. AND Hay, David L. AND Battle. 2012. “Ross, Macdonald, and a Theory for the Dynamics and Control of Mosquito-Transmitted Pathogens.” <em>PLOS Pathogens</em> 8 (4): 1–13. <a href="https://doi.org/10.1371/journal.ppat.1002588">https://doi.org/10.1371/journal.ppat.1002588</a>.</p>
</div>
<div id="ref-volterra">
<p>Volterra, Vito. 1926. “Variazioni E Fluttuazioni Del Numero d’individui in Specie Animali Conviventi.” <em>Memoria Della Reale Accademia Nazionale Dei Lincei. Ser. VI, Vol. 2.</em></p>
</div>
</div>
</div>

</div>
<div class="main">
<p><strong>SemanticModels.jl: A Framework for Automatic Composition of Scientific Models Across Domains</strong></p>
</div>
<div class="main_bottom">
<img id="main-img-left" src=figs/semanticmodels.png>
<img id="main-img-center" src=>
<img id="main-img-right" src=figs/gtri.png>
</div>
</div>



</body>
</html>

BIN
conference-poster/poster.pdf View File


+ 127
- 0
conference-poster/poster.rmd View File

@@ -0,0 +1,127 @@
---
main_topsize: 0.2 #percent coverage of the poster
main_bottomsize: 0.1
column_numbers: 3
#ESSENTIALS
title: ''
title_textsize: "90pt"
author:
- name: '**Micah Halter**'
affil: 1
main: true
email: Micah.Halter@gtri.gatech.edu
- name: Kun Cao
affil: 1
main: false
email: 'Kun.Cao@gtri.gatech.edu'
- name: James Fairbanks
affil: 1
main: false
email: 'James.Fairbanks@gtri.gatech.edu'
affiliation:
- num: 1
address: Georgia Tech Research Institute
primary_colour: "#101B49"
secondary_colour: "#101B49"
main_fontfamily: "Rasa"
main_findings:
- '**SemanticModels.jl: A Framework for Automatic Composition of Scientific Models Across Domains**'
logoleft_name: 'figs/semanticmodels.png'
logoright_name: 'figs/gtri.png'
<!-- logocenter_name: https&#58;//raw.githubusercontent.com/brentthorne/posterdown/master/images/qr-code-black.png -->
output:
posterdown::posterdown_betterport:
self_contained: false
pandoc_args: --mathjax
number_sections: false
bibliography: refs.bib
link-citations: true
---

<style>
.main p {
margin-left: 0em;
}
#main-img-left {
width: 35%;
bottom: 1.25in;
}
#main-img-right {
width: 30%;
bottom: 1.25in;
}
</style>

```{r, include=FALSE}
knitr::opts_chunk$set(echo = FALSE,
warning = FALSE,
tidy = FALSE,
message = FALSE,
fig.align = 'center',
out.width = "100%")
options(knitr.table.format = "html")
```

# Introduction

Scientific progress comes from adapting and extending models from prior work to address new problems. The task of modifying existing source code and performing semantically sound model fusion can be difficult as the models scale, and become more complex. We propose `SemanticModels.jl`, a category theory-based framework for defining meta-modeling tasks and semantic information extraction. Our objectives are threefold:

1. Convert scientific models to a universal wiring diagram representation [@fong];
1. Use this representation to augment and compose these models at domain-level semantics;
1. Generate executable code and visualize results.

# Methods

We leveraged techniques from static and dynamic program analysis to process executable versions of scientific models to extract the mathematical representation of the Lotka-Volterra Predator-Prey [@volterra] and the MacDonald Ross Disease Model [@ross]. Next we performed meta-modeling tasks defined as domain specific transformations to augment and compose the two models. Finally, we can visualize the new model as well as generate executable code and visualize the simulation's results.

# Results

Figures \@ref(fig:volterra) and \@ref(fig:ross) show the universal, domain specific wiring diagrams that domain scientists can manipulate to create new models. Figure \@ref(fig:ross) also illustrates the amount of math that is contained in the simple Malaria wiring diagram.
These wiring diagrams allow us to perform model augmentation and fusion at the semantic level, and the fusion of the differential equations is handled automatically.
We then augmented the Lotka-Volterra model from Figure \@ref(fig:volterra) to model mutual predation and composed it with the Ross Malaria model from Figure \@ref(fig:ross) to get a model that simulates both, shown in Figure \@ref(fig:volterraross).
Finally we generated the code to simulate the new model, and the results are shown in Figures \@ref(fig:results) and \@ref(fig:resultszoom).

```{r, volterradiffeq, fig.cap='The full set of differential equations of the Lotka-Volterra model that we want to represent as a mathematical structure grounded in semantic knowledge.', out.width="80%"}
knitr::include_graphics("figs/volterra-diffeq.png")
```

```{r, volterra, fig.cap='**Top**: Lotka-Volterra predator-prey model represented as a semantically rich wiring diagram;<br />**Bottom**: Lotka-Volterra Predator-Prey Model represented as a structured mathematical object', out.width="95%"}
knitr::include_graphics("figs/lotka-volterra.png")
```

<br />

```{r, ross, fig.cap='**Left**: Ross Malaria model represented as a wiring diagram;<br />**Right**: The full set of differential equations represented by the wiring diagram on the left.', fig.show="hold", out.width="49%"}
knitr::include_graphics(c("figs/malaria.png", "figs/malaria-diffeq.png"))
```

<br />

```{r, volterraross, fig.cap='**Top**: The generated Lotka-Volterra and Malaria disease spread model as a wiring diagram;<br />**Bottom**: The generated Lotka-Volterra and Malaria disease spread model as a mathematical structure.', out.width="90%"}
knitr::include_graphics("figs/volterra-ross.png")
```

```{r, results, fig.cap='The results of the mutual predation and Malaria model showing the predator-prey interaction between the mosquitoes and birds (purple and yellow lines respectively)', out.width="95%"}
knitr::include_graphics("figs/result_graph.png")
```

<br />

```{r, resultszoom, fig.cap='Zooming into the beginning of the graph in Figure \\@ref(fig:results) shows the initial role Malaria had on the system.', out.width="95%"}
knitr::include_graphics("figs/result_graph_zoom.png")
```

<br />

# Conclusions

1. `SemanticModels.jl` provides a framework for scientists to easily augment and transform existing scientific models.
1. Using a universal representation allows for models to be easily composed across scientific domains.
1. New models can automatically generate executable code and produce verifiable results.
1. Our representation can be adapted to curate more types of models that can then be used in the same workflow.

# Acknowledgments

This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Agreement No. HR00111990008.

# References

+ 27
- 0
conference-poster/refs.bib View File

@@ -0,0 +1,27 @@
@article{volterra,
title = {Variazioni e fluttuazioni del numero d’individui in specie animali conviventi},
author = {Volterra, Vito},
year = {1926},
journal = {Memoria della Reale Accademia Nazionale dei Lincei. Ser. VI, vol. 2.}
}

@misc{fong,
author = {Brendan Fong},
title = {Seven Sketches in Compositionality: An Invitation to Applied Category Theory},
year = {2018},
eprint = {arXiv:1803.05316},
}
@article{ross,
author = {Smith, David L. AND Battle, Katherine E. AND Hay, Simon I. AND Barker, Christopher M. AND Scott, Thomas W. AND McKenzie, F. Ellis},
journal = {PLOS Pathogens},
publisher = {Public Library of Science},
title = {Ross, Macdonald, and a Theory for the Dynamics and Control of Mosquito-Transmitted Pathogens},
year = {2012},
month = {04},
volume = {8},
url = {https://doi.org/10.1371/journal.ppat.1002588},
pages = {1-13},
abstract = {Ronald Ross and George Macdonald are credited with developing a mathematical model of mosquito-borne pathogen transmission. A systematic historical review suggests that several mathematicians and scientists contributed to development of the Ross-Macdonald model over a period of 70 years. Ross developed two different mathematical models, Macdonald a third, and various “Ross-Macdonald” mathematical models exist. Ross-Macdonald models are best defined by a consensus set of assumptions. The mathematical model is just one part of a theory for the dynamics and control of mosquito-transmitted pathogens that also includes epidemiological and entomological concepts and metrics for measuring transmission. All the basic elements of the theory had fallen into place by the end of the Global Malaria Eradication Programme (GMEP, 1955–1969) with the concept of vectorial capacity, methods for measuring key components of transmission by mosquitoes, and a quantitative theory of vector control. The Ross-Macdonald theory has since played a central role in development of research on mosquito-borne pathogen transmission and the development of strategies for mosquito-borne disease prevention.},
number = {4},
doi = {10.1371/journal.ppat.1002588}
}

+ 31
- 0
letterhead/README.md View File

@@ -0,0 +1,31 @@
# GTRI-Letterhead

This is a LaTeX and pandoc markdown template for the Georgia Tech Research Institute Letterhead.

## Usage

LaTeX template:

`latexmk --pdf letter.tex`

pandoc markdown template:

`pandoc --template letter_template.tex letter.md -o letter.pdf`

pandoc template options:

| option | details | required |
|-------------|---------------------------|--------------------------------------------|
| `name` | sender name | required |
| `company` | sender company | required |
| `position` | sender position | required |
| `address` | sender street address | required |
| `citystate` | sender city, state zip | required |
| `phone` | sender phone number | optional |
| `email` | sender email | optional |
| `website` | sender website | optional |
| `recipient` | receiver name and address | required |
| `date` | date of letter | optional, default today |
| `opening` | letter greeting | optional, default "To Whom It May Concern" |
| `closing` | letter closing | optional, default "Sincerely" |
| `options` | `logo` or `no-logo` | optional, default `logo` |

gtri-ltr.sty → letterhead/gtri-ltr.sty View File


letter.md → letterhead/letter.md View File


letter.pdf → letterhead/letter.pdf View File


letter.tex → letterhead/letter.tex View File


letter_template.tex → letterhead/letter_template.tex View File


logo.png → letterhead/logo.png View File


Loading…
Cancel
Save