-
Weekly Meetings will be held to understand the progress of each student
-
Weekly progress report send to: Abhishek Sivaram (as5397@columbia.edu), Resmi Suresh (rsm2189@columbia.edu), and cc Prof. Venkat (venkat@columbia.edu)
-
End of the semester expected report and presentation (Due 12/15)
-
Robustness and Efficiency, network entropy translates to Robustness/Efficiency of networks
-
We aim to understand this parallel between Efficiency and Robustness, and Entropy of networks – by applying this to certain biological systems.
-
Genetic algorithm -- analytical framework for predicting different structures given the relative importance of efficiency and robustness
-
Genetic algorithm - We are also looking at scale-up of the complex networks work.
- Read papers mentioned
- Search for networked datasets from biological and chemical examples
-
It is expected that the student take sample datasets from http://deeplearning.net/datasets/ (from each subcategory), trains them and does analysis based on their trained model
-
Sensitivity study on the input layer -- Change the input layer node values to see how the output behaves. This will tell you the important nodes as well.
-
The weight distribution was hypothesized to follow a power law distribution, which shall be studied as well.