I worked as a research assistant in this project. Through engagement and collaboration with stakeholders and academic/industry experts, I applied text mining to build a portal for AD research to create a central resource and knowledge base for technology identification and training.
This was my first time using Magnetic Resonance Imaging (MRI) data for both demented and nondemented adults to build classifiers that predicts whether a subject will be diagnosed to develop dementia. Two datasets are used: one deals with cross-sectional MRI data for adults aged between 18 to 96, and the other deals with longitudinal MRI data for older adults between 60 to 96.
This is my final project of the CIS 545 - Big Data Analysis in the Spring 2022 semester. We worked as a team to tackle this air pollution big data problem as air is what keeps humans alive. Monitoring it and understanding its quality is of immense importance to our well-being. At the end, we also won the Best Visualization outstanding project among the whole class.
The goals in this study are 1) to identify important words associated with positive ratings and negative ratings, and 2) to predict ratings using different methods.
Predicting students’ academic performance at school using regression methods is not a new area of interest. Machine learning methods, however, are relatively new in this field and it has been flourishing in recent years. According to Ghorbani and Ghousi (2020), due to technological advancements, predicting students’ performance at school is among the most beneficial and significant research topics nowadays. Therefore, we believe that it is a meaningful area for us to focus on and we decide to analyze the Open University Learning Analytics Dataset to study the student’s academic performance.
This course prepares me to use tools from those fields effectively in applied contexts and build skills including (1) feature representations of spreadsheet-based or text datasets; (2) training classification and regression models for prediction tasks; (3) evaluation of machine learning model accuracy and error analysis; and (4) reasoning about predictive models and making tradeoffs like bias vs. variance, granularity and annotation complexity in labeled training data, and the ethical application of predictive modeling to human-centered data.
This course prepares me to use tools from those fields effectively in applied contexts and build skills including (1) feature representations of spreadsheet-based or text datasets; (2) training classification and regression models for prediction tasks; (3) evaluation of machine learning model accuracy and error analysis; and (4) reasoning about predictive models and making tradeoffs like bias vs. variance, granularity and annotation complexity in labeled training data, and the ethical application of predictive modeling to human-centered data.