# 2nd research iteration results & roadmapping ###### tags: `gitcoin` `Reports` :::info Updated by December 2021 ::: *Authors: Danilo Lessa Bernardineli* --- ## Results under the original scope --- - The definitions & methodology of the research plan were sucessfully implemented - This includes creating rewiring optimizers based on Hill Climbing and Simulated Annealing that were pushed to the NetworkX module --- <sup><sub>*Source: https://github.com/gitcoinco/gitcoin_cadcad_model/tree/main/optimality_gap*</sub></sup> ![](https://i.imgur.com/oLryk5R.png) --- - We have the capacity of knowing the Optimality Gap under limited circumstances - The current capacity is limited due to the complexity factor of the current QF algorithm implementation, which is quadratic: $O(n^2)$ --- <sup><sub>*Source: https://github.com/gitcoinco/gitcoin_cadcad_model/blob/main/optimality_gap.ipynb* ![](https://i.imgur.com/BnNOGqM.png)</sub></sup> --- - The further testing of the proposed hypothesis is blocked by further R&D - Large-scale analysis and simulations requires potentially tens of thousands of evaluations, or the adoption of better heuristics. A more comprehensive mathematical description of the algorithm could also create shortcuts. --- <sup><sub>*Source: https://github.com/gitcoinco/gitcoin_cadcad_model/blob/main/optimality_gap.ipynb* ![](https://i.imgur.com/jNJ1XPL.png)</sub></sup> --- ## Results Outside the scope --- - We've written drafts of mathematical specs that could result in drastic improvements to QF evaluation performance on operations and research. --- <sup><sub>*Source: https://hackmd.io/7HYPPp7zQcShQLrKQ2KOxQ?view#Simple-Quadratic-Match*</sub></sup> ![](https://i.imgur.com/ELASsRS.png) --- - We've tested the effect of attack vectors under limited subsets of the Gitcoin Grants data. --- <sup><sub>*Source: https://github.com/jiajia20/GitCoin_attack/blob/main/attack_vector.ipynb*</sub></sup> ![](https://i.imgur.com/zcy1bKL.png) --- <sup><sub>*Source: https://github.com/gitcoinco/gitcoin_cadcad_model/blob/main/attack_vector_ab_test.ipynb*</sub></sup> ![](https://i.imgur.com/tsewgPN.png) --- - We've promoted public research interest on Gitcoin and Quadratic Funding. --- <sup><sub>*Source: https://www.notion.so/Gitcoin-Modelling-Co-Lab-TL-DR-dba6e25863a1413a81e75f989f4a1f67*</sub></sup> ![](https://i.imgur.com/PJVUHFG.png) --- ## Next Steps --- - Improving the scalability of the Optimality Gap calculations - Starting point heuristics based on attack vectors - Iterative rewiring algorithm for QF - Linear Algebra formalism for QF - Researching analytical solutions for optimizing QF --- - Investigating the effect of attack vectors under a range of scenarios on synthetic / subset scenarios - What if the distribution of the amounts change? --- - Searching for attack vector patterns on synthetic / subset scenarios
{"metaMigratedAt":"2023-06-15T19:45:10.756Z","metaMigratedFrom":"Content","title":"2nd research iteration results & roadmapping","breaks":true,"contributors":"[{\"id\":\"a8ef47b3-fccc-4c1e-86b7-c5b392de83ca\",\"add\":3010,\"del\":76}]"}
    1989 views
   owned this note