# Computational Public Policy: Scope

This page provides additional details on the kinds of public policy problems that motivate the graduate program.

## Problem domain

### Scope

* monetary policy
* banking and financial sector management
* agriculture and natural resource management
* forecasting aggregates and prices
* access to energy, housing and water
* public infrastructure and logistics
* forecasting distributions
* trade
* economic impacts of climate change
* managing demographic change

### Requirements

* data science and visualization
* economic management
* data collection and data management
* machine learning and estimation for economic data
* optimization and logistics

### Target skills

* simulation and mathematical modeling
* linear programming
* mixed integer and quadratic programming
* network methods
* software engineering and programming
* high performance computing
* CPU and GPU based parallelization
* supervised and unsupervised learning

## Methodology

* Tools, not solutions
* Provide tools, training, code, knowledge and techniques
* Trainees lead in developing solutions
* Instructors learn and then specialize their training

* Common problems, shared solutions
* Shared knowledge, code and data
* Enhanced collaboration

* Open source code for computational methods
* Avoid commerical licences (Gurobi is $20K and support is extra)
* Training in open source techniques enhances independence
* Open development improves robustness and facilitates collaboration

* Promotes long run skill acquisition
* Build human capital paired with open source tools
* Teaching tools tailored to a domain facilitates learning

* Flexible skills and flexible tools
* Enhance discovering of information
* Provide capacity to answer new questions

* Access to new data
* Skills to access to new data that can be used for problem solving

Note: Machine learning techniques to extract useful development data from novel places such as video / text analysis etc. [This project](https://centerforcollectivelearning.org/urbanperception), for example, used google street view data to map the "quality of an urban environment".

## Output and Activities

* Training workshops (online and in-person)
* Specialized lecture series (written and video)
* Specialized code
* Complementary adaptive textbook
1. tailor content to their interests
2. power a community for exploring new and novel topics at a global scale
3. provide a platform to provide guest lecturing