Regression & Classification Algorithms
Using real-world datasets on patients and Virginia schools as case studies, I've constructed clinical models that generate dozens of informative visuals, display correlations, delineate key indicators, and predict values based on machine learning algorithms. These projects properly implement the data science lifecycle: exploratory data analysis, filtering noise, redundant, and duplicate variables, selecting predictors, then generating the models.
The models I created include predicting whether patients will be readmitted to a hospital and the standard score for schools given certain indicators. The csv files are available in the Tableau section.
Natural Language Processing (NLP) & Sentiment Analysis on Amazon Reviews
Analyzing a series of reviews on Amazon about certain products, my program uses keywords to identify the overall sentiment with negative, neutral, and positive ratings.
Unsupervised ML, Anomaly Detection, & Clustering Algorithms on Zillow Homes
This specific algorithm uses the libraries pycaret and pandas to first identify anomalies with the knn model, then filtering them out to create run a clustering algorithm.
HTML, CSS, JavaScript, & API Integration
Enjoy a game of Tic-tac-toe.
A stylish weather website that integrates an API to display the weather at locations around the world.
A simple app to search through food products on a company website.
Java, C++
Click the images to see the code run in a YouTube video.
Divergent Shunting
This versatile C++ program uses a shunting yard algorithm to convert expressions to different formats (like Infix, Postfix, etc) and creates binary trees out of them.
Business Trip
This Java program efficiently filters from a standardized list of flights containing departure and arrival locations and times in the form of a text file.
Superb Scheduling
This program reads the classes of students in a data file and generates all the possible permutations of a matrix whose dimensions are determined by the number of periods and classes per period. Then, it creates the best schedule using chi-squared algorithms to minimize variation in the number of students per period and the number of scheduling conflicts where students are enrolled in more than one class in a single period.