Nicolas Kiryczun

Link to the project's GitHub repository

The Pareto Principle suggests that 20% of the input drives 80% of the results. In MMA, that 20% consists of the sport’s superstars. This project identifies these prospects using historical data from UFC Fighters (2010 – 2024).

Traits such as arm length, winning streaks, and finishing rates were used to compare superstars with average UFC fighters. Weighting the different traits based on their association with superstars, the “Superstar Score” was created. This function intakes raw fighter info and outputs a singular score used to determine a fighters superstar potential.

Visualizations were critical in defining the story of a superstar. Side-by-side boxplots helped compare superstars with average UFC fighters. Scatterplots allowed for correlations to be found or disproven between variables, and bar graphs allowed for more detailed analysis on individuals leading the pack in certain traits.

Tools

Python  •  Pandas  •  Matplotlib  •  Seaborn

Python Skills Used

Data loading  •  Data wrangling  •  Defining functions  •  For-loops   •  If-else statements  •  Feature engineering  •  DataFrame/Series manipulation  •  Matplotlib plot creation   •  Seaborn plot manipulation

Statistical Skills Used

Descriptive statistics  •  Comparative analysis  •  Correlation visualization  •  Trend visualization

Link to the dashboard on Tableau Public

Link to the project's GitHub repository

This project's first phase aimed at using SQL to clean, transform, and analyze global population data to find the out fastest growing countries/regions in the world. This raw dataset contained yearly populations for every country beginning in 10,000BC. The initial cleanup process involved filtering out all non-countries and removing years with inconsistent population data. Next, obtaining the yearly growth rate for each year of every country led to the conclusion that the Arabian Peninsula currently has some of the fastest growing countries in the world.

This takeaway was used as the inspiration for creating a Tableau dashboard covering this region of the world. Transformed datasets were uploaded to Tableau to be datapoints for different KPI's, visualizations, and maps. These metrics all highlight the difference between the Arabian Peninsula's growth rate compared to the rest of the world. This dashboard is interactive and allows the user to filter by year.

SQL Skills Used

Data cleaning  •  Data transformation  •   VIEW statements  •   JOIN clauses  •   Subqueries  •   Window functions

Statistical Skills Used

Descriptive statistics (Mean, Standard Deviation)  •   Compound Growth Analysis

Tableau Skills Used

Interactive filters  •  Geographic Visualization  •   Data Grouping  •   Time-Series Analysis

Link to the project's GitHub repository

This project allowed me to turn raw numbers into an actionable format, then use it to display the data’s visual story. The first half of this project consisted of this transformation process, where I used SQL to group daily index prices into yearly averages. I then organized these numbers into a table fit for visualization. I connected my database with Power BI and began choosing the visual format in which these numbers could be shown. The result is a dashboard which gives a glimpse into the S&P 500’s expansive history, along with insights into some of the most famous US market crashes/bull runs. This dashboard encourages interaction and creativity. The full, interactive dashboard is available upon request.

SQL Skills Used

VIEW statements  •  CASE expressions  •  JOIN clauses   •   Common Table Expressions (CTEs)  •   GROUP BY clauses   •  Aggregate Functions   •  Subqueries

Power BI Skills Used

Interactive Slicers  •  Dynamic Visualizations  •  Conditional Formatting   •  Data Grouping  •  KPI Cards  •  Time-Based Filtering