Research Project

For the research project, you will:

1. Research a current topic in Artificial Intelligence.
   
   Use papers from the following conferences in your research:
   • AAAI Conference on Artificial Intelligence (AAAI)
     • Call for Papers from AAAI 2013
     • Papers from AAAI 2012
     • Papers from AAAI 2011
     • Papers from AAAI 2010
   • International Joint Conference on Artificial Intelligence (IJCAI)
     • Call for Papers from IJCAI 2013
     • Papers from IJCAI 2011
     • Papers from IJCAI 2009
     • Papers from IJCAI 2007
   • International Conference on Machine Learning (ICML)
     • Papers from ICML 2013 (Under Construction)
     • Papers from ICML 2012
     • Papers from ICML 2011
     • Papers from ICML 2010
   • Or select a conference from the Wiki for Calls For Papers (WikiCFP) and track down their conference papers.
   
   NOTE: Unless you are "special," you will initially have little to no clue about what most research papers are discussing ... and may feel deterred. No worries. We all start there. Just keep researching until it starts to make sense ... i.e. keep adding vectors to that high dimensional space until your innate unsupervised learning algorithms have enough to go on to automatically detect the "important" features, start classifying vectors of those features, and eventually generate new feature vectors.
   
   Acceptable reseach topics include:
   • Deep Learning
   • Constraints, Satisfiability, and Search
   • Knowledge-Based Information Systems
   • Knowledge Representation, Reasoning, and Logic
   • Machine Learning
   • Multiagent Systems
   • Natural Language Processing
   • Reasoning about Plans, Processes and Actions
   • Reasoning Under Uncertainty
   • Robotics
   • Simultaneous Localization and Mapping (SLAM)
   • Computer Vision
   • Cognitive Systems
   • Computational Sustainability and Artificial Intelligence
   • Planning and Scheduling
   • Web and Data Mining
   • Web and Knowledge-Based Information Systems
   • Compressed Sensing
   • Reinforcement Learning
   • Social Networks
   • Topic Modeling
   • Feature Learning
   • Spectral Learning and Tensors
   • Online Learning
   • General Support Vector Machines and Decision Tree Methods
   • Dimensionality Reduction and Semi-Supervised Learning
   • Robust Statistical Methods for Regression or Classification
   • Nearest Neighbor and Metric Learning for Regression or Classification
   • Bayesian Methods for Regression or Classification
   • Transfer Learning
   • Statistical Learning and Inference
   • Clustering
   • Crowd Sourcing and Large Scale Learning
   • ... and the list goes on ... and on ... and on ...
   
   
2. Based on the research that you've done, do one of the following:
   • Come up with an original idea that improves on this research (i.e. show that your idea improves on an existing algorithm by comparing results on datasets).
   • Critique the research that you've done (i.e. advantages and disadvantages of various algorithms with comparisons on datasets).
   
   
3. Write a working protoype in original C++ code related to your chosen research topic.
   
   Note: Use g++ version 4.7.2 as your compiler. This compiler is available on linprog4.cs.fsu.edu as g++47.
   
   Note: You may also use the C++11 Standard Library, Open Source Computer Vision (OpenCV) version 2.4.4, or Boost version 1.53.0 (esp. the Boost Graph Library (BGL)) in your code. These libraries are available on linprog4.cs.fsu.edu at the following locations:
   • C++11 Standard Library: Directly available in g++47
   • Open Source Computer Vision (OpenCV) version 2.4.4: /usr/local/lib64/
   • Boost version 1.53.0: /usr/local/include/boost_1.53.0/
   
   Note: OpenCV and certain parts of Boost uses shared libraries; therefore, use -L/usr/local/lib64 -Wl,-rpath,/usr/local/lib64 in your makefile.
   
   
4. Write a Conference-Style Research Paper (a maximum of eight pages) referencing the results of that working prototype.
   
   Choose one of the following conference formats:
   • AAAI Conference on Artificial Intelligence (AAAI)
   • International Joint Conference on Artificial Intelligence (IJCAI)
   • International Conference on Machine Learning (ICML)
   
   This Conference-Style Research Paper will have sections in the following order:
   1. Your Title
   2. Your Contact Information
   3. Abstract
   4. Introduction
   5. Background / Prior Work
   6. Your Work
   7. Method of Evaluation / Experiment
   8. Your Results
   9. Conclusion & Future Work
   10. (Optional) Acknowledgements
   11. References
   
   Note: You must provide and use at least five references in your Research Paper.
   
   
5. Upload your Conference-Style Research Paper, your working C++ protoype (only .cpp files and .h/.hpp files), your makefile, and links to your datasets to the Research Project assignment in BlackBoard:
   1. Logon to the BlackBoard course site at https://campus.fsu.edu.
   2. Once at the course site, press the "Project" button on the left-hand-side menu.
   3. Select "Research Project" from the available assignments.
   4. In the Submission text box, provide the links to your datasets.
   5. Attach PDF file version of your Conference-Style Research Paper.
   6. Attach a ZIP file of your working C++ protoype (only .cpp files and .h/.hpp files) and makefile.
   7. Submit your Research Project.
   Note: Your working C++ prototype must be able to compile on linprog4.cs.fsu.edu using your makefile.