DATA SCIENCE OVERVIEW

2020

 RACIAL AND ECONOMIC IMPACTS OF COVID-19 IN KENYA 

  • Aimed to assess the economic and racial impacts on infections and deaths due to COVID-19 in Kenya. The paper will use both quantitative and qualitative research designs and methods to describe the extent of the pandemic 

2020

DESIGNING A DEEP LEARNING ALGORITHM THAT PREDICTS THE NUMBER CASES AND DEATHS OF THE COVID-19 PANDEMIC

  • Predicted the number of deaths due to the COVID-19 pandemic based on latitude, longitude, and the regional cumulative number of confirmed cases provided by the Johns Hopkins University Medical School. Used a random forest algorithm and a bootstrap method to build a model and then tested for accuracy.

2020

DEVELOPING AN ALGORITHM FOR PRIVATE AUTOMATED CONTACT TRACING (PACT) OF COVID-19

  • Helped research as part of the MIT Lincoln Laboratory's 2020 PiPact Project

  • Built a Raspberry Pi based Bluetooth signal collection platform using Python and Git and developed processing algorithms that perform the proximity detection

  • Studied whether and how automated exposure detection can provide measurable improvements in manual contact tracing efforts to slow infection rates

2020

ANALYZING OPEN-SOURCE DATA TO GENERATE ACTIONABLE INTELLIGENCE FOR DISASTER RESPONSE

  • Helped research under the guidance of Professor Jeff Liu Ph.D. of the MIT Lincoln Laboratory

  • Extracted and classified features from the raw Light Detection and Ranging (LiDAR) open-source data of the MIT Lincoln Laboratory via machine learning techniques and data products for decision-makers.

  • Analyzed Geospatial Data using GeoPANDAS and geospatial information systems.

  • Processed satellite images and analysis using multispectral imaging

  • Git Hub Repo: https://github.com/bwsi-remote-sensing-2020

2019

DESIGNING A NOVEL METHOD FOR PERSONALIZING RECOMMENDATIONS TO DECREASE PLASTIC POLLUTION

  • Research published on the Cornell arXis journal

  • Individually conducted various statistical and computational analysis and developed an application using Google App Scripts that will search for the percent mismanaged plastic waste of the user's inputted country from a large database, determine the country's standing, and output a personalized set of recommendations.

2019

DESIGNING AN ALGORITHM THAT DETECTS FAKE AMAZON REVIEWS

  • Designed an algorithm in Java that detects “fake” reviews on Amazon using semantic analysis such as looking for exaggerated words or similar reviews that may indicate that the review is fake.

  • Tested and analyzed the accuracy of the results as well as interpreting it from the aspect of the six qualities of code.

2018

DETERMINING THE OPTIMAL FONT FOR MAXIMUM CONCENTRATION FROM THE AUDIENCE DURING VARIOUS PRESENTATION

  • Determined the optimal font for various presentations by measuring students' ability to efficiently read texts in different fonts when it is formatted in blocks or in bullet-points.

  • Tested on more than 200 Monta Vista High School students randomly with single-bias and analyzed the data; and drew a conclusion with statistical analysis and graphs.

 

 RACIAL AND ECONOMIC IMPACTS OF COVID-19 IN KENYA 

The main aim of this paper is to assess the economic and racial impacts on infections and deaths due to COVID-19 in Kenya. The paper will use both quantitative and qualitative research designs and methods to describe the extent of the pandemic. Currently, the impact and spread of infections and increase in dearth cases are feared to be exacerbated by the huge number of people living in poverty; a weak health infrastructure; overcrowding in informal settlements; and poor access to basic services such as clean water, sanitation, and hygiene. However, the country has been encountered by massive disruption of the economic activities in terms of GDP decline, massive job losses, and racial discriminations of different Kenyan citizens in different countries which is not a solution to the problem but a catalyst to infections and deaths. While the country had started experiencing the economic and racial impacts before and after the COVID-19, the rapid spread of the virus has put Kenya into economic and racial troubles. COVID -19 has been greatly impacted by the Kenyan economy and racial lines negatively as indicated in financial markets, disruption of global supply chains, a reversal of prior monetary and fiscal policies, volatility of Kenyan currency, and reduction in Diasporas remittances. Truly, COVID-19 has immensely affected the Kenyan economic and racial spheres.

 

DESIGNING A DEEP LEARNING ALGORITHM THAT PREDICTS THE NUMBER CASES AND DEATHS OF THE COVID-19 PANDEMIC

2020

With the rapidly growing COVID-19 pandemic, public health experts and decision-makers are faced with a multi-criteria problem where it is necessary to reconcile public health objectives, social well-being, and economic performances. While non-targeted lockdown strategies have proven their epidemiological effectiveness in controlling the spread of the virus, these strategies are non-sustainable solutions for cities’ economies because urban mobility is important for a city’s livability and economic productivity.  From my Exploratory Data Analysis, I concluded that local transmission of COVID-19 severe. Under the hypothesis that a recurring pattern between the number of confirmed cases and deaths exists, the number of confirmed cases was set as a variable for the model that predicts the number of deaths. The model was tested and compared with the actual COVID-19 data to calculate its accuracy.

 

Screen Recording of EDA

 

Screen Recording of Model

 

DESIGNING A NOVEL METHOD FOR PERSONALIZING RECOMMENDATIONS TO DECREASE PLASTIC POLLUTION

2020

According to Our World in Data, Third World countries (i.e. India) tend to have a higher share of plastic waste that is inadequately managed while First World countries (i.e. The US) have higher plastic waste generation per person. This difference in the characteristics of plastic pollution depending on the country's standing results in varying optimal recommendations for users depending on which country they live in. Through Big Text and OSOME meme analysis, I constructed a list with optimal recommendations for First World and Third World countries. Based on the list, I designed a User Interface with Google Apps Scripts that provide personalized recommendations based on the country’s standing and user’s preferred difficulty and reassessed the code based on the six qualities of code. The UI's purpose is to aid people who wish to help solve plastic pollution by offering a set of personalized tasks for each user and keeping their progress accountable through a point tracking system. With a significant number of users, the application could eventually contribute to solving plastic pollution.

 
 

DESIGNING AN ALGORITHM THAT DETECTS FAKE AMAZON REVIEWS

2019

Often, there are suspicious amazon reviews that seem to be excessively positive or have been created through a repeating algorithm. I moved to detect fake reviews on Amazon through semantic analysis in conjunction with metadata such as time, word choice, and the user who posted. I first came up with several instances that may indicate a review isn't genuine and constructed what the algorithm would look like. Then I coded the algorithm and tested the accuracy of it using statistical analysis and also analyzed it based on the six qualities of code.

 

DETERMINING THE OPTIMAL FONT FOR MAXIMUM CONCENTRATION FROM THE AUDIENCE DURING VARIOUS PRESENTATIONS

2018

There are many factors that come into play during presentations such as the speaker's tone, delivery skills, posture, slide designs, and more. However, I thought that the font style of presentations also plays a big role in delivering the presentation's content. I tested more than 200 Monta Vista High School students on their ability to efficiently read texts in different fonts when formatted in blocks or in bullet-points and drew a conclusion with statistical analysis and graphs.