Repository designed for the course "Design Thinking and Predictive Analytics for Data Products" at the University of California San Diego.
Here, you will find the notebooks from the course lectures, which provide introductory and practical examples of Machine Learning applications in projects. Additionally, you can explore my Final Project, where I developed two machine learning models from scratch: one for regression and another for classification.
For this challenge, I am developing a regression model to predict cars' fuel consumption in kilometers per liter. The dataset can be found in the following GitHub repository: Link to repository.
Here I am working on a classification model that aims to predict expense categories, based on previously labeled bank transaction data. This model will be implemented in my job, so I will be using a fictitious dataset.
The problem I am endeavoring to solve revolves around enhancing both my own and my colleagues' performance by automating the manual task of classifying bank transactions daily. We handle financial data from multiple clients, providing them with essential information about their expenses so they can exercise complete control over their monthly financial planning. The objective of this model is to streamline and optimize our workflow, alleviating the burden of mundane tasks and freeing up valuable time to focus on providing even better services to our clients. Through the implementation of this advanced solution, we aim to elevate service standards, boost productivity, and empower our clients to make informed decisions with confidence in their financial stability.
I have trained bank transactions data from the past 6 months. The labels are classifying expenses into the following categories:
