Giuseppe Celentano's Portfolio

Showcasing Projects & Achievments

This website is maintained by Giuseppe Celentano

Projects

My Projects

Prudential Financial Convert C# Tool to Python


I was a member of the Convert C# Tool to Python team at Prudential Financial. We were given the source code to a legacy tool being used at Prudential and our team was tasked with converting it to Python. This was part of a larger effort to convert Prudential's legacy tools to Python to update their programs. We were able to reduce the code size by around 70%, reduced the run time of the program, and create a simple and more maintainable program. The project was part of the New Jersey Institute of Technology capstone program and lasted about five months. There was a capstone project showcase after students completed their industry sponsored projects and our team achieved first place in the showcase.

Bitcoin (BTC) Price Predictor


Bitcoin Price Predictor
Bitcoin Price Predictor written in Python that takes the closing price of BTC per day from the past few years to predict a specific month's closing price on each day. The graph included is the real price for the test month versus the predicted price. There is also an experimental version of the predictor that uses Amazon DeepAR instead of regression. View Code

Artificial Intelligence Snake Python Game



 Untrained Snake AI                             Trained Snake AI Using DQN Learning

This AI program was written in Python and uses deep learning to learn how to play the game. It works by assigning a reward value and a penalty value to certain actions, like 10pts for eating the food, and -10pts for crashing into itself or the wall. The AI then attempts to maximize this score by making actions based on certain states. States include danger-left, danger-right, danger-front, and the same for food. When first starting, the snake scores randomly and usually dies within a few seconds, but after about 20-30 minutes of training the snake can reach scores of about 20-35pts. View Code

Ethereum (ETH) Price Predictor


Ethereum Price Predictor
Ethereum Price Predictor written in Python. The graph included is the real price for the test month versus the predicted price. There is also an experimental version of the predictor that uses Amazon DeepAR instead of regression. This project was implemented using the same strategy as my bitcoin price predictor but you can see it is less accurate because Ethereum is considered more volatile than Bitcoin and Ethereum is newer causing less stable data. View Code

Social Media Music Website




This is a website that was a collaboration between two NJIT students and myself. It is a social media page focused on talking about and sharing music and music related topics. It was designed to be used for a smaller target audience like a college campus, a high school, a club/organization, or any other small to medium sized group. If you would like to explore the website further, contact me and I can give you a set of login credentials. The site remains private to prevent tampering. Features of the website include online/offline instant messaging, rich-text and image posts, comments, admin functions such as creating users and blocking content, Spotify integration, view accounts' Spotify favorites, Soundcloud and Youtube Music integrations, and profile search. This is a working prototype that is hosted on Amazon Web Services (AWS). View Code

Brick-Breaker Java Game




A Brick-Breaker Game Made Using Java. Score to win is 450pts and you only get one life. View Code

More Resources

My personal Github profile linked below shows more of my projects. View my profile to see the code behind these projects or to see more of my projects. My Github

Linked below is a public Google Drive folder that includes various data science projects I have completed including the BitCoin Price Predictor shown above. All of the code is written in Python and they run similarly to Jupyter Notebook. Google Colab Folder