CS 6702: Topics in Computational Sustainability

Coursework: Reaction Paper

 

Computational Sustainability is an emerging and multidisciplinary field that aims to identify and address new computational problems. This assignment is intended to give the student exposure to computational problems sustainability researchers and policy‐makers encounter in their work and any existing research on these problems. To accomplish this goal, the student will have to identify the links between research papers from conferences and journals in Computer Science and the relevant sustainability fields or describe a sustainability problem he/she has encountered. A non‐exhaustive list of relevant journals and conferences is included at the end of this document as well as a lits of computational sustainability papers from previous conferences.  

 

Students may complete a reaction paper, a research problem description, or an annotated bibliography to fulfill this assignment. Guidelines for each follow. Students are encouraged to choose a topic for the reaction assignment that matches their final project interests.

 

Students are expected to give a short presentation in class about this assignment.

 

Students are encouraged to consult the faculty team and course assistants to discuss ideas.

 

Reaction Paper

Deliverables: Paper (approximately 2-3 pages)

 

A reaction paper should identify one or two computational issues concerning a sustainability topic, and then react to current research papers that encounter or address these issues. A reaction paper can be written in four steps:

 

1.     Find a seed paper that addresses or encounters computational issue(s) in a sustainability topic. The seed paper should be a strong starting point from which to build a ÒnetworkÓ of papers you can react to. Here are some ways to find a seed paper:

a.     Use a paper whose presentation you have seen in class or in a related talk.

b.     Use one of the background readings from the course website.

c.     Search through key journals in the area you are considering (see the list at the end of this document).

d.     Use Google Scholar. Try combining your favorite sustainability topics with computational keywords. For example, combining terms like poverty, ecosystem, agriculture, renewable energy, or conservation with keywords like algorithm, network, machine learning, modeling, game theory, or optimization.

e.     Ask the faculty/research team and course assistants for advice. Come to office hours!

2.     Populate a list of research papers to serve as the references for your reaction paper. Here are some ways to populate your list of papers:

a.     Check the papers cited by your seed paper.

b.     Use the ÒCited ByÓ link on Google Scholar to find papers that cite the seed paper. Other bibliographic databases like CiteSeer may also help.

c.     Check the authorÕs website for related work.

d.     Search for the paper on Google. Is the paper discussed on forums or blogs that point to other sources?

3.     React. Using the references compiled above, you should present your sustainability topic and the computational issue(s) encountered in this sustainability topic. You should then form a reaction to the current research. Good reactions will do the following:

a.     Address some of the following questions about each reference:

                                               i.     What is (are) the main problems the author is addressing?

                                             ii.     Why is the problem important for sustainability?

                                            iii.     What is the central claim, argument, or point of the paper?

                                            iv.     What assumptions does the paper make?

                                             v.     Are the models or techniques presented in the paper supported by theory, experiment, and/or evidence?

                                            vi.     What are the main strengths and weaknesses of the paper?

                                          vii.     What future work can come from this paper?

b.     Consider the references collectively.

                                               i.     How do the papers relate to one another?

                                             ii.     What is the overall picture they portray?

                                            iii.     Is there a next logic step?

4.     Re-edit your work. No one gets it exactly right the first time around.

 

 

Research Problem Description

 

Deliverables: Paper (approximately 2-3 pages)

 

 

A research problem description should identify a specific computational problem encountered in one or more sustainability fields, and then a survey of the current research that addresses the problem. A research problem presentation is a good option for students currently engaged in sustainability‐related research. A good research problem presentation will do the following:

 

1.     Identify and define the computational problem and its sustainability applications.

2.     Identify the current best algorithms, models and techniques for solving the problem.

3.     Present the problem and current best solutions. Your presentation should explain and motivate your chosen problem to the class. You should then present the research you identified in Step 2, addressing some of the following issues:

a.     What weaknesses or assumptions exist in the current research?

b.     What computational requirements (time, space) must problem solutions address? How does current research address them?

c.     Are there other computational methods that could be used to solve the problem that current research has not explored?

d.     Do existing problem formulations / models adequately address the underlying sustainability issues?

e.     Are there optimizations/improvements that could be made to current techniques that are themselves research questions?

 

Annotated Bibliography

Deliverables: 20 references in bibtex format and one-­‐paragraph annotations for each reference.

 

The annotated bibliography should be composed of references that address a topic in Computational Sustainability of the studentÕs choice. Students should choose topics with a computational focus that apply to sustainability. Some example topics are:

1.     Optimization methods for species conservation.

2.     The modeling of complex adaptive systems like ecosystems, the ocean‐ atmosphere system, systems from epidemiology, social networks, etc.

3.     Models and methods from economics and game theory for the management of natural resources.

4.     Computational methods for crisis management in underdeveloped or at‐risk regions.

5.     Human computation overview (potential sustainability applications)

6.     Games with a purpose overview (sustainability applications: e.g., Foldit)

7.     Crowdsourcing and citizen science overview (potential sustainability applications: e.g., Christmas Bird Count, Mushroom Obsrever, and Merlin http://www.allaboutbirds.org/labs/)

8.     Design concepts and techniques with applications to sustainability

 

A one‐paragraph annotation should be provided for each reference. Annotations should be a short review of the paper with a focus on the computational problems and solutions found therein.

 

Some Relevant Journals/Conferences

 

 

AAAI-11 Computational Sustainability Special Track Papers

 

á      Green Driver: AI in a Microcosm 
Jim Apple, Paul Chang, Aran Clauson, Heidi Dixon, Hiba Fakhoury, Matthew L. Ginsberg, Erin Keenan, Alex Leighton, Kevin Scavezze, Bryan Smith

á      Enforcing Liveness in Autonomous Traffic Management 
Tsz-Chiu Au, Neda Shahidi, Peter Stone

á      Policy Gradient Planning for Environmental Decision Making with Existing Simulators 
Mark Crowley, David Poole

á      Dynamic Resource Allocation in Conservation Planning 
Daniel Golovin, Andreas Krause, Beth Gardner, Sarah J. Converse, Steve Morey

á      Water Conservation Through Facilitation on Residential Landscapes 
Rhonda Hoenigman, Elizabeth Bradley, Nichole Barger

á      Incorporating Boosted Regression Trees into Ecological Latent Variable Models 
Rebecca A. Hutchinson, Li-Ping Liu, Thomas G. Dietterich

á      A Large-Scale Study on Predicting and Contextualizing Building Energy Usage 
J. Zico Kolter, Joseph Ferreira

á      The Steiner Multigraph Problem: Wildlife Corridor Design for Multiple Species 
Katherine J. Lai, Carla P. Gomes, Michael K. Schwartz, Kevin S. McKelvey, David E. Calkin, Claire A. Montgomery

á      Hybrid Planning with Temporally Extended Goals for Sustainable Ocean Observing 
Hui Li, Brian Williams

á      Stochastic Model Predictive Controller for the Integration of Building Use and Temperature Regulation 
Alie El-Din Mady, Gregory Provan, Conor Ryan, Kenneth Brown

á      Linear Dynamic Programs for Resource Management 
Marek Petrik, Shlomo Zilberstein

á      Logistic Methods for Resource Selection Functions and Presence-Only Species Distribution Models 
Steven Phillips, Jane Elith

á      Modeling and Monitoring Crop Disease in Developing Countries 
John Alexander Quinn, Kevin Leyton-Brown, Ernest Mwebaze

á      Learned Behaviors of Multiple Autonomous Agents in Smart Grid Markets 
Prashant P. Reddy, Manuela M. Veloso

á      Efficient Energy-Optimal Routing for Electric Vehicles 
Martin Sachenbacher, Martin Leucker, Andreas Artmeier, Julian Haselmayr

á      Verifying Intervention Policies to Counter Infection Propagation over Networks: A Model Checking Approach 
Ganesh Ram Santhanam, Yuly Suvorov, Samik Basu, Vasant Honavar

á      Discovering Life Cycle Assessment Trees from Impact Factor Databases 
Naren Sundaravaradan, Debprakash Patnaik, Naren Ramakrishnan, Manish Marwah, Amip Shah

á      Decentralised Control of Micro-Storage in the Smart Grid 
Thomas Voice, Perukrishnen Vytelingum, Sarvapali Ramchurn, Alex Rogers, Nicholas Jennings

 

AAAI-12 Computational Sustainability Special Track Papers

á      The Automated Vacuum Waste Collection Optimization Problem 
Ram—n BŽjar, CŽsar Fern‡ndez, Carles Mateu, Felip Manyˆ, Francina Sole-Mauri, David Vidal

á      MOMDPs: A Solution for Modelling Adaptive Management Problems 
Iadine Chades, Josie Carwardine, Tara G. Martin, Samuel Nicol, Regis Sabbadin, Olivier Buffet

á      Fine-Grained Photovoltaic Output Prediction Using a Bayesian Ensemble 
Prithwish Chakraborty, Manish Marwah, Martin Arlitt, Naren Ramakrishnan

á      A Novel and Scalable Spatio-Temporal Technique for Ocean Eddy Monitoring 
James H. Faghmous, Yashu Chamber, Shyam Boriah, Frode Vikeb¿, Stefan Liess, Michel dos Santos Mesquita, Vipin Kumar

á      Learning Non-Stationary Space-Time Models for Environmental Monitoring 
Sahil Garg, Amarjeet Singh, Fabio Ramos

á      Patrol Strategies to Maximize Pristine Forest Area 
Matthew Paul Johnson, Fei Fang, Milind Tambe

á      Pre-Symptomatic Prediction of Plant Drought Stress Using Dirichlet-Aggregation Regression on Hyperspectral Images 
Kristian Kersting, Zhao Xu, Mirwaes Wahabzada, Christian Bauckhage, Christian Thurau, Christoph Ršmer, Agim Ballvora, Uwe Rascher, Jen Leon, Lutz PlŸmer

á      Lagrangian Relaxation Techniques for Scalable Spatial Conservation Planning 
Akshat Kumar, Xiaojian Wu, Shlomo Zilberstein

á      An Intelligent Battery Controller Using Bias-Corrected Q-learning 
Donghun Lee, Warren B Powell

á      Sensing the Air We Breathe — The OpenSense Zurich Dataset  
Jason Jingshi Li, Boi Faltings, Olga Saukh, David Hasenfratz, Jan Beutel

á      Sustaining Economic Exploitation of Complex Ecosystems in Computational Models of Coupled Human-Natural Networks 
Neo D. Martinez, Perrine Tonnin, Barbara Bauer, Rosalyn C. Rael, Rahul Singh, Sangyuk Yoon, Ilmi Yoon, Jennifer A. Dunne

á      Global Climate Model Tracking Using Geospatial Neighborhoods 
Scott McQuade, Claire Monteleoni

á      Coupling Spatiotemporal Disease Modeling with Diagnosis 
Martin Gordon Mubangizi, Caterine Ikae, Athina Spiliopoulou, John A. Quinn

á      Prediction and Fault Detection of Environmental Signals with Uncharacterised Faults 
Michael Alan Osborne, Roman Garnett, Kevin Swersky, Nando de Freitas

á      Non-Intrusive Load Monitoring Using Prior Models of General Appliance Types 
Oliver Parson, Siddhartha Ghosh, Mark Weal, Alex Rogers

á      Factored Models for Multiscale Decision-Making in Smart Grid Customers 
Prashant P. Reddy, Manuela M. Veloso

á      Cooperative Virtual Power Plant Formation Using Scoring Rules 
Valentin Robu, Ramachandra Kota, Georgios Chalkiadakis, Alex Rogers, Nicholas R. Jennings

á      Robust Cuts Over Time: Combatting the Spread of Invasive Species with Unreliable Biological Control 
Gwen Spencer

á      Improving Hybrid Vehicle Fuel Efficiency Using Inverse Reinforcement Learning 
Adam Vogel, Deepak Ramachandran, Rakesh Gupta, Antoine Raux

á      Scheduling Conservation Designs via Network Cascade Optimization 
Shan Xue, Alan Fern, Daniel Sheldon

á      An Efficient Simulation-Based Approach to Ambulance Fleet Allocation and Dynamic Redeployment 
Yisong Yue, Lavanya Marla, Ramayya Krishnan

 

CompSustÕ09 Papers

http://www.computational-sustainability.org/compsust09/schedule.php

CompSustÕ10 Papers

http://www.computational-sustainability.org/compsust10/schedulewithvideos.php

CompSustÕ12 Papers:

http://www.computational-sustainability.org/compsust12/master.php

Other Relevant Journals/Conferences

 

 

á      Journal of Environmental Economics and Management (JEEM)

á      Conservation Biology

á      Biological Conservation

á      Ecology

á      Ecology Applications

á      Ecology Economics

á      PNAS Sustainability Science (special issues)

á      IEEE Spectrum

á      Resource and Energy Economics

á      Environmental and Resource Economics

á      Science

á      American Journal of Agricultural Economics

á      Energy Policy

á      AAAI

á      AAMAS (International Conference on Autonomous Agents and Multiagent Systems)

á      ECAI (European Conference on Artificial Intelligence)

á      IJCAI (International Joint Conference on Artificial Intelligence

á      UAI (Uncertainty in Artificial Intelligence)

á      ICML (International Conference on Machine Learning)

á      KDD (Conference on Knowledge Discovery and Data Mining)

á      CP (International Conference on Principles and Practices of Constraint Programming)

á      CPAIOR (International Conference on Integration of Artificial Intelligence and Operations Research Techniques in Constraint Programming for Combinatorial Optimization Problems)

á      CHI - Computer Human Interaction

á      UbiComp (HUC) - Ubiquitous Computing/Handheld and Ubiquitous Computing

á      HCI – Human Computer Interaction

á      CSCW - Conference on Computer Supported Cooperative Work