Welcome to My Portfolio!

Overview

Tools and Technologies: Python pandas NumPy Tableau LangChain Streamlit

This project analyzes the nutritional composition of Canadian foods using The Canadian Nutrient File dataset. Through data preprocessing, k-means clustering, and regression analysis, it standardizes nutrient values, classifies food subcategories by protein-to-fat ratios, and evaluates nutrient density per calorie. Insights are presented via interactive Tableau dashboards, while a LangChain-powered AI chatbot enables structured data queries, providing actionable insights for informed dietary choices.

Overview

Tools and Technologies: Python pandas NumPy matplotlib seaborn scikit-learn XGBoost

A credit card company provided a dataset containing demographic and financial data for 30,000 account holders over the past six months. Each row represents a credit account, with a label indicating whether the account holder defaulted (failed to make the minimum payment) in the month following the six months.

This project aims to develop a predictive model that forecasts credit card defaults based on demographic and financial data. Key tasks include loading and exploring the dataset using pandas, creating visualizations with matplotlib and seaborn, implementing machine learning models with scikit-learn and XGBoost, and applying regularization techniques to reduce overfitting. The project also evaluates model performance, interprets predictions, and delivers actionable business recommendations to improve risk management strategies.

Overview

Tools and Technologies: Python pandas NumPy matplotlib seaborn statsmodels scikit-learn

This project explores the relationship between a country's population size and GDP per capita and its success in the 2024 Olympic Games. By analyzing these variables, we aim to understand how they influence the total number of medals won by different countries. The study also involves clustering countries based on their Olympic performance and socioeconomic context, providing insights into patterns and trends affecting their athletic achievements. Through this analysis, we seek to uncover the extent to which economic and demographic factors contribute to Olympic success.

In this project, we began with data preprocessing to clean and transform the dataset, ensuring it was ready for analysis. We then conducted a correlation analysis to explore the relationships between population, GDP per capita, and the total number of medals won, finding moderate positive correlations. Moving on to regression analysis, we built models to quantify how well these socioeconomic factors predict medal counts, revealing that neither population nor GDP per capita alone is a strong predictor. We performed clustering and PCA analysis to gain deeper insights, allowing us to group countries based on their Olympic performance and visualize these clusters effectively. These steps collectively led us to conclude that while population and GDP per capita have some influence, other factors, such as sports infrastructure and government support, play significant roles in determining Olympic success.

Overview

Tools and Technologies: Python pandas NumPy matplotlib seaborn TensorFlow Keras scikit-learn

This project leverages a sequential neural network model to help a bank identify customers most likely to accept personal loan offers. By analyzing demographic and financial data—such as income, credit card spending, education level, and CD account ownership—the model provides actionable insights to optimize marketing efforts and resource allocation, increasing loan acceptance rates and driving business growth.

The model achieved 98% accuracy, 96% precision, and 87% recall for identifying loan acceptors, ensuring reliability and effectiveness. Key visualizations, including loss/accuracy plots and a confusion matrix, confirm the model's ability to generalize and identify patterns in customer behaviour, enabling more targeted outreach.

Overview

Tools and Technologies: Python LangChain OpenAI API Streamlit SQLAlchemy pandas

This project aims to build AI agents that interact with databases using natural language. The agents simplify the querying and extracting insights from various datasets, enhancing the efficiency of data-driven decision-making.

Overview

Tools and Technologies: Python pomegranate PyTorch NetworkX NumPy matplotlib

The Wumpus World is a classic problem in artificial intelligence that involves designing an agent to navigate a grid-based environment filled with hazards, such as pits and a fearsome Wumpus while searching for gold. The agent must operate under uncertainty, leveraging logical reasoning and perception to make decisions in a partially observable world.

This project focuses on implementing an intelligent Wumpus World agent capable of:

Perceiving environmental cues such as breezes (near pits) and stenches (near the Wumpus).
Applying logical inference to deduce the safe path to the gold.
Balancing exploration and caution to avoid fatal encounters.

The project highlights key AI concepts, including decision-making under uncertainty, logical reasoning, and search algorithms, providing a practical application of intelligent agent design in a challenging and dynamic environment.

Overview

Tools and Technologies: Python pandas streamlit

This Streamlit app provides insights into YouTube channel performance, tracking subscriber growth, views, watch hours, likes, comments, and shares. It supports daily, weekly, monthly, and quarterly trends with interactive charts and thoughtful insights, allowing users to analyze engagement, visualize trends, and filter data by custom date ranges.

Overview

Tools and Technologies: Python pandas NumPy matplotlib seaborn

The project aims to analyze the quality and characteristics of red and white wines by addressing the following questions:

How do the average values of features differ between high-quality and low-quality wines for both red and white wine types?

How does the alcohol content distribution vary between high-quality and low-quality wines for red and white wine types?

What are the differences in total sulfur dioxide content distribution between high-quality and low-quality wines for red and white wine types?

Overview

Tools and Technologies: R ggplot2 GGally olsrr mctest

The project aims to create a regression model to predict the Critical Temperature (Tc) of different superconducting materials.

Guiding questions:

What factors are the most accurate for predicting a given material's Critical Temperature (Tc)?

Can we use the measured properties of a given material to predict its Critical Temperature (Tc)?

Overview

Tools and Technologies: R dplyr tidyr ggplot2

This project has thoroughly examined the factors contributing to variations in life expectancy across different countries, reinforcing its status as a crucial indicator of population well-being and the effectiveness of health systems. Through exploratory data analysis, we have gained valuable insights into how health, socio-economic, and demographic factors impact life expectancy.

The findings reveal significant disparities between developed and developing nations. Countries with lower life expectancy often grapple with substantial challenges, such as elevated adult and child mortality rates and a higher prevalence of diseases like HIV/AIDS. These elements demonstrate strong negative correlations with life expectancy, emphasizing the urgent need for targeted public health interventions. Our analysis underscores the pivotal role of vaccination coverage in enhancing health outcomes. Higher vaccination rates for diseases like Hepatitis B, Polio, and DTP3 are strongly associated with improved life expectancy, underscoring the importance of robust public health measures in the fight against infectious diseases.

Our analysis revealed that the size of a population has little correlation with life expectancy, suggesting that demographic factors alone do not determine longevity. However, it did show a positive correlation between healthcare expenditure—both as a percentage of total expenditure and GDP—and life expectancy. This finding underscores the critical need for sustained investment in healthcare systems to improve health outcomes.

Furthermore, socio-economic factors such as the Human Development Index and years of schooling emerged as more robust predictors of life expectancy than GDP alone. This indicates that holistic development, encompassing education and human welfare, is essential for improving life expectancy.

In conclusion, enhancing life expectancy is a complex endeavour requiring a multifaceted approach. This includes strategies to reduce mortality rates, increase vaccination coverage, and invest substantially in healthcare and socio-economic development. By addressing these interconnected factors, we can make significant strides toward improving population well-being and ensuring healthier futures for all.

Overview

Tools and Technologies: Tableau

The British Airways Customer Review Analytics Dashboard provides a detailed analysis of customer reviews from March 2016 to October 2023. It showcases key metrics such as overall satisfaction, cabin crew service, in-flight entertainment, food and beverage quality, and seat comfort. Users can explore trends in customer sentiment over time, compare ratings by seat type and travel class, and analyze average review scores across different countries. With interactive filters for date, traveller type, seat class, and region, this dashboard is a powerful tool for understanding customer feedback and identifying areas for service improvement.

Overview

Tools and Technologies: Tableau

The Sales & Profit Analytics Dashboard offers an interactive analysis of sales and profit trends, allowing users to explore key performance metrics through customizable filters. Users can dynamically filter data by year, product category, subcategory, region, state, and city to gain insights into sales performance across different locations and product segments. This dashboard enables businesses to track trends, identify top-performing products and regions, and optimize sales strategies. With its user-friendly filters, the dashboard serves as a valuable tool for data-driven decision-making and performance analysis.

Overview

This article explores the relationship between a country's population size and GDP per capita and its success in the 2024 Olympic Games. By analyzing these variables, we aim to understand how they influence the total number of medals won by different countries. The article also involves clustering countries based on their Olympic performance and socioeconomic context, providing insights into patterns and trends affecting their athletic achievements. Through this analysis, we seek to uncover the extent to which economic and demographic factors contribute to Olympic success.

Overview

In hypothesis testing, we start with the null hypothesis (H₀), assuming no effect or difference. We then collect our observed results—the actual data from the study—and calculate the p-value, which tells us how surprising these results are if H₀ is true. A small p-value (e.g., < 0.05) leads us to reject H₀, considering the alternative hypothesis (H₁) as a possible explanation. A large p-value suggests our results align with H₀, so we fail to reject it. Importantly, the p-value is not the probability that H₀ is true; instead, it’s the probability of observing our results, or more extreme ones, under the assumption that H₀ is correct.