Audience & Trend Insights
Predicting Income, Informing Policy
Using machine learning to forecast income levels and guide smarter policy, business, and social impact decisions, earning second place in a global data competition.
Year :
2023
Industry :
Public Policy & Government Analytics
Client :
Self: Personal Project
Project Duration :
3 weeks
The Challenge:
Accurately tracking income levels between census years is a global challenge. Traditional data collection methods are costly, time-consuming, and often outdated by the time results are available. Yet, understanding income distribution is critical for shaping policy, targeting services, and addressing inequality.
This project aimed to build a machine learning model capable of predicting whether an individual earns above or below $50,000 based on demographic, employment, and socio-economic variables, using a large-scale dataset of 20th-century U.S. population data.
Insight:
Income is not just a number, it’s a reflection of education, employment, geography, and life circumstances. By analysing patterns across these variables, we can forecast income levels with high accuracy, providing decision-makers with real-time, actionable insights that would otherwise require years of traditional data gathering.
Approach:
Data Scope – Analysed ~200,000 training records and ~100,000 test records containing demographic, employment, and socio-economic attributes.
Feature Selection: Key variables included age, education, occupation, marital status, industry, hours worked, gains/losses, and migration history.
Model Development: Built and tested machine learning models to classify individuals as earning above or below the $50,000 threshold.
Ethical Considerations: Addressed potential biases, ensured transparency, and considered privacy implications in model design.
Approach:
The model demonstrated applications in public policy, finance, marketing, HR, and healthcare, from forecasting tax revenue to targeting social programs, while highlighting the importance of ethical AI in socio-economic decision-making.
I achieved second place in the competition, garnering industry attention and raising awareness of how predictive analytics can transform both business strategy and policy development.
More Projects
Audience & Trend Insights
Predicting Income, Informing Policy
Using machine learning to forecast income levels and guide smarter policy, business, and social impact decisions, earning second place in a global data competition.
Year :
2023
Industry :
Public Policy & Government Analytics
Client :
Self: Personal Project
Project Duration :
3 weeks
The Challenge:
Accurately tracking income levels between census years is a global challenge. Traditional data collection methods are costly, time-consuming, and often outdated by the time results are available. Yet, understanding income distribution is critical for shaping policy, targeting services, and addressing inequality.
This project aimed to build a machine learning model capable of predicting whether an individual earns above or below $50,000 based on demographic, employment, and socio-economic variables, using a large-scale dataset of 20th-century U.S. population data.
Insight:
Income is not just a number, it’s a reflection of education, employment, geography, and life circumstances. By analysing patterns across these variables, we can forecast income levels with high accuracy, providing decision-makers with real-time, actionable insights that would otherwise require years of traditional data gathering.
Approach:
Data Scope – Analysed ~200,000 training records and ~100,000 test records containing demographic, employment, and socio-economic attributes.
Feature Selection: Key variables included age, education, occupation, marital status, industry, hours worked, gains/losses, and migration history.
Model Development: Built and tested machine learning models to classify individuals as earning above or below the $50,000 threshold.
Ethical Considerations: Addressed potential biases, ensured transparency, and considered privacy implications in model design.
Approach:
The model demonstrated applications in public policy, finance, marketing, HR, and healthcare, from forecasting tax revenue to targeting social programs, while highlighting the importance of ethical AI in socio-economic decision-making.
I achieved second place in the competition, garnering industry attention and raising awareness of how predictive analytics can transform both business strategy and policy development.
More Projects
Audience & Trend Insights
Predicting Income, Informing Policy
Using machine learning to forecast income levels and guide smarter policy, business, and social impact decisions, earning second place in a global data competition.
Year :
2023
Industry :
Public Policy & Government Analytics
Client :
Self: Personal Project
Project Duration :
3 weeks
The Challenge:
Accurately tracking income levels between census years is a global challenge. Traditional data collection methods are costly, time-consuming, and often outdated by the time results are available. Yet, understanding income distribution is critical for shaping policy, targeting services, and addressing inequality.
This project aimed to build a machine learning model capable of predicting whether an individual earns above or below $50,000 based on demographic, employment, and socio-economic variables, using a large-scale dataset of 20th-century U.S. population data.
Insight:
Income is not just a number, it’s a reflection of education, employment, geography, and life circumstances. By analysing patterns across these variables, we can forecast income levels with high accuracy, providing decision-makers with real-time, actionable insights that would otherwise require years of traditional data gathering.
Approach:
Data Scope – Analysed ~200,000 training records and ~100,000 test records containing demographic, employment, and socio-economic attributes.
Feature Selection: Key variables included age, education, occupation, marital status, industry, hours worked, gains/losses, and migration history.
Model Development: Built and tested machine learning models to classify individuals as earning above or below the $50,000 threshold.
Ethical Considerations: Addressed potential biases, ensured transparency, and considered privacy implications in model design.
Approach:
The model demonstrated applications in public policy, finance, marketing, HR, and healthcare, from forecasting tax revenue to targeting social programs, while highlighting the importance of ethical AI in socio-economic decision-making.
I achieved second place in the competition, garnering industry attention and raising awareness of how predictive analytics can transform both business strategy and policy development.