A Reinforcement Learning/AI project that maps the problem of error correction in quantum computers to a board game. Topological error-correcting codes are one of the most promising routes for realizing large-scale, fault-tolerant quantum computers. One of the biggest hurdles for realizing quantum computers is their sensitivity to noise from the environment and from measurements. In this project, we implement techniques from Reinforcement Learning to model error-correction for various noise models in quantum computation. We implement this tech- nique on a popular error-correcting code known as the surface code.
This project was commissioned by the World Wildlife Fund (WWF) and the Coca Cola Foundation. The goal of the project was to come up with mathematical models and to quantitatively measure the affects of reforestation and various other projects by the Coca Cola foundation's social responiosibility program to become water neutral. The work was published and presented at the "International water Sustainability conference" held at UET, Lahore.
Abstract: This research aimed to study the affect deforestation on soil infiltration rates and effective porosity in an effort to gauge the sensitivity of the soils of Ayubia National Park to deforestation. Furthermore, we formulated a mathematical model that related soil infiltration rates and water retention capacities of the soil to the vegetation (trees, shrub and grass cover) while accounting for variables such as slope, soil texture, canopy cover etc through multi-variable linear regression. A total of 18 sites, ten from forested areas and 8 from deforested areas were documented and these two data sets were compared. The sites were in located in the larger Namli Mera District. Lastly, the Coca Cola Company as part of its comprehensive social responsibility program, wants to become water neutral. In addition to other endeavors, it has funded WWF in its reforestation program in the areas surrounding Ayubia National Park. The company wished to know how much water it will save by reforesting several target locations, one of which is also located in Namli Mera. We have attempted to provide a framework for calculating this value.
In this project, we model the spread of Hepatitis C as a multiple compartment mathematical model in terms of various parameters and simulate the results using MATLAB.
Abstract: Hepatitis C is an infectious disease that harms the liver. It is caused by the hepatitis C virus and the infection leads to adverse aff ects like damage of the liver and ultimately to cirrhosis. In our model we would be considering two modes of transport which are statistically significant. One is the transmission through intravenous drug use and the other mode is through blood transfusions. We would have separate compartments for both of these transmission modes. So the susceptible compartment can go into either Acute 1 or Acute 2 compartment depending on the 1st mode of transmission. It depends on the infection rate of susceptible which in turn depends on the contact rate etc. You can either recover from the acute condition or progress to the chronic infection compartment. From the chronic compartment you can be moved to Quarantine or you could recover. If you recover you can go back to the susceptible compartment. If you are Quarantined then some fraction moves back to susceptible and some goes to the acute infection compartment. There is a recruitment rate at which people keep getting added to the susceptible population. All compartments have a natural death rate, however Acute, Chronic and Quarantine have also their own corresponding death rates in addition to the natural death rate. Also there are similar compartments Acute 2, Chronic 2, Quarantined 2 and Recovered 2 for the other mode of transmission too.