Data Science and Analytics Projects
In this project, we used natural language processing techniques to analyze the live chat from the Law&Crime network's YouTube channel.
This work is in process. Our goal is to understand the sentiment of the audience towards those involved in the lawsuit.
This project involved designing and delivering end-to-end business intelligence solutions for the company by creating a PowerBI sales report for the company.
​
On this project, I identified and tracked KPIs like (Net Sales, Profit, Orders, and Returns) and showed insights on a monthly, weekly, quarterly, and yearly basis. This also involved collecting the data, cleaning, transforming the data, and creating new custom tables and columns. Building table relationships, and in page navigation.
Stock price prediction with images
This project involved time series analysis of Angilent Technologies Inc. stock price and elaborating a neural network to predict the stock price using images of the stock line graph as input in our model. We were able to demonstrate that stock price forecasts based on images can generate very accurate predictions, similar to or better than traditional methods.
This project involved analyzing the movie's information and using natural language preprocessing (NLP) techniques and word embedding models to design a movie recommendation app to show the most similar movies.
​
If a movie is not in the DataFrame, the app dynamically pulls information from IMDb and TMDb APIs, performing in-depth calculations to suggest the most similar movies and storing the new movie in a database to later add it to the DataFrame weekly.​
This project involved evaluating various sampling methods to determine the most effective approach for obtaining a representative sample of our target population, thereby reducing data collection costs and time. Additionally, I delved into the data to gain insights into its distribution, enhancing our understanding of key characteristics within the target population. This combined effort ensures that our collected data accurately reflects the population and informs decision-making processes to optimize engagement strategies.
This project involved the exploration of Divvy bike users' data to identify distinct customer segments and their unique behaviors. Our goal was to provide insights to create marketing campaigns tailored to optimize engagement with Divvy riders, ultimately increasing annual memberships. By using tools such as Tableau, we were able to create engaging visualizations that allowed stakeholders to explore customer segmentation data in real time.
