loves the Earth.
colors.png

Data Science

Data Science

Rochelle has worked with data ranging from patients on experimental cancer therapy trials at Weill Cornell Medical College to studying econometrics in graduate school to her more recent work with ESG (environmental, social, governance) data and sustainability analysis. Rochelle looks forward to future explorations with Python, machine learning techniques, mining open data and data visualization. 


 Cluster analysis for indicator for undernourishment, maternal mortality rate and access to internet use vs. ICT development for 83 countries (from left to right)   

Cluster analysis for indicator for undernourishment, maternal mortality rate and access to internet use vs. ICT development for 83 countries (from left to right)

 

Measuring the effect of ICT indicators on sustainable development indicators using regression and cluster analysis 

Building off Rochelle's work for Huawei's ICT Sustainable Development Goals Benchmark, Rochelle gathered data from the Sustainable Development Solutions Network, ITU and the World Bank to test the relationship between ICT (information communications technology) performance and sustainable development of 83 countries.

She ran a linear regression, and then Lasso and Ridge regressions to isolate relevant features. These features she then used in a cluster analysis, an unsupervised machine learning technique, to determine if there were similarities among  countries.  

She found that health-related sustainable development indicators are most correlated with ICT development, suggesting an area where ICT investment has proven successful for health outcomes historically and should continue to be focused to further benefit health-related outcomes, particularly in developing countries. 

Full presentation and Python notebook can be accessed on Github.