Hi, I'm Dennis!

Unraveling the Dynamics of Airfare Price Predictions

Exploratory Data Analysis, Classification & Regression Models, Neural Networks, Model Evaluation

A project addressing the challenges of airfare price unpredictability and limited filtering capabilities in flight aggregator apps. By utilizing machine learning, the final model enables users to predict how ticket prices vary based on selected amenities, offering deeper insights into the connection between airfare costs and specific flight features.

Check it out

AirBnb Reservation Predictions in Seattle

Amazon Simple Storage Service (S3), MySQL, Data Pipeline

A project examining the fluctuation of Airbnb listings in Seattle, with a core focus on constructing a data pipeline incorporating Amazon S3 and MySQL. The cloud-based storage was successfully established and machine learning algorithms were applied to the archived dataset, determining Random Forest yielded the best predictive capabilities.

Check it out

Predictive Analysis of Car Trade-In Values

Exploratory Data Analysis, Classification & Regression Models, Neural Networks, Model Evaluation

This data science project aimed to develop a predictive model for estimating a car’s resale value based on its features. Using a dataset sourced from CarMax, a leading used car retailer, the analysis explored the relationship between vehicle attributes and trade-in values to generate accurate resale potential forecasts.

Check it out

Predictive Analysis of Loan Approvals

Cloud Computing, AWS Sagemaker, XGBoost

A project examining the creditworthiness of bank clientele through the implementation of classification models. Utilizes exploratory data analysis to gauge underlying issues, such as class imbalance, within the dataset and construct a predictive model by optimizing and finetuning XGBoost’s hyperparameters.

Check it out

Airline Flight Delays

CART, C5.0, Random Forest, and Neural Networks

This study analyzed over 500,000 flight records to explore how attributes like airline, departure/arrival locations, flight duration, and day of the week relate to travel delays. Using classification models—CART, C5.0, Random Forest, and Neural Networks—the study assessed performance through accuracy, sensitivity, specificity, precision, and F1 score.

Check it out

Predicting Airline Customer Satisfaction

Feature Importance, AIC Score, Backward Elimination, Data Wrangling

While delays and cancellations—like those linked to Southwest—often impact passenger satisfaction, other factors such as Wi-Fi, food and drinks, cleanliness, and seat comfort also influence customer experience. This study aims to assess how significantly these amenities affect overall customer satisfaction.

Check it out

Natural Gas Forecasting

Time Series Analysis, ACF Plots, Exponential Smoothing, ARIMA

As a key energy source, natural gas plays a vital role in electricity generation and global infrastructure. Utilizing time series forecasting and decades of natural gas price data, this study investigates monthly price fluctuations using six different forecasting methods; aiming to identify patterns and improve accuracy of future price predictions.

Check it out

Airline Review Text Classification and Topic Modeling

Web Scraping, Text Classification, Topic Modeling, Tokenization

This project involved extracting airline reviews from TripAdvisor and SkyTrax using the Playwright library. Text classification and topic modeling were applied to the review content alone and then enhanced by incorporating airline attribute ratings to improve model performance and insight.

Check it out