MY PROJECT PORTFOLIO

DATA VISUALIZATION

Visualizing COVID-19

Since first being recorded late December 2019 in China, COVID-19 has spread around the world and been declared a pandemic by the World Health Organization. Because of this, billions of persons have been sent into lockdown as health services struggle to cope. As of June 14th, 2020, the estimated total worldwide COVID-19 cases reached 11,240,173 and data show that the number of confirmed cases within 6 countries account 50% of worldwide cases. It was also found that despite the lockdown and social distancing measures put in place, cases within the United States continue to increase at an alarming rate and can be seen as concerning since states have begun easing their lockdown restrictions.

RECESSION PREDICTION

Evaluating Statistical Machine Learning Algorithms in Economic Recession Prediction (May 2020)

Machine learning has grown to become an effective way of classifying observations into groups and making predictions on these groupings by finding patterns within data based on a set of shared characteristics. Because of this, the use of machine learning algorithms has grown rapidly within various fields.

Within this study, I used empirical data and statistical models to evaluate whether there exist patterns in macroeconomic data that can be used to classify different types of economic periods and whether these patterns can be used to improve the accuracy of recession predictions. I found that the pattern recognition models developed within this study identified features within the data which accurately separated recessions from non-recessions. I also found that these features predicted [the probability of] recessions with a high level of accuracy. I compared the results derived from my models to those of the Federal Reserve's predicted probabilities of a recession, and I concluded that the use of statistical machine learning techniques significantly improves the accuracy of recession predictions.

ITEM CLASSIFICATION

Machine Learning for Environmental Researchers: Plant Classification (December 2019)

In machine learning, classification is a supervised learning concept which basically categorizes a set of data into classes. Classification algorithms are used when the desired output is a discrete label; they're helpful when the answer to what is being investigated falls under a finite set of possible outcomes.

​

In this study, I classify types of physical trees based on their characteristics within the data set. This study is relevant because trees contribute to the environment by providing oxygen, preserving soil, providing an ecosystem for wild animals and reducing the intensity of the greenhouse effect. It is also known that a tree’s chance of survival depends on factors such as climate, exposure to sunlight, location and soil type. If it is possible to classify trees based on observable patterns within the data, then it might assist arborists determine characteristics of particular ecosystems.

REGRESSION ANALYSIS

A study on the effect that the Human Development Index has on a country’s Gross Domestic Product per capita (May 2019)

This study attempts to examine the effect that a country’s Human Development Index (HDI) has on it’s Gross Domestic Produce per capita (GDPPC). More specifically, the study should determine how well the HDI is a predictor of the standard of living within an economy.

In this study, I used empirical data and regression models to determine the effect that HDI has on GDPPC, while accounting for different factors. I found that there exists a non-linear relationship between GDPPC and HDI, where model coefficients are most significant when determining the percentage change in GDPPC, and a unit change in HDI is associated with a 9% increase in GDPPC (holding the other variables within the model constant).

CRITICAL PATH ANALYSIS

Examining the Efficiency of Various Bus Routes, SGU Grenada (March 2017)

The St George’s University (SGU) is Grenada’s only university. Apart from that, the university is mainly an accredited medical, business and information technology school. Noting that these are major occupational fields in the world, it is obvious that the school is flooded with individuals commuting on daily, or even hourly, basis.

​

The university has established a transportation system which has made commuting easier for students. The main issues, however, arise when it comes to the time between the arrivals and departures of buses and the amount of time spent on particular routes. This study investigates the use of alternative routes and critical path analysis to determine whether the issues stated can be mitigated by using a different route. 

​

​