ML Minor - Senior Projects

Sample of available research projects (updated April 2018).

Contact your favorite professor or for updates.



Leman Akoglu

Interactive visualization and summarization of social circles in online social networks

Profiling users by inferring missing attributes from their friends'

Nina Balcan

Privacy Preserving Machine Learning

Life Long Learning

Kai-Min Chang

Exploiting Longitudinal EEG Input in a Reading Tutor

Anind Dey &
Jennifer Mankoff

Machine learning for large-scale mobile personal and crowd data mining and visual/predictive analytics.

Dynamic feature ordering for interactive predictions.

Artur Dubrawski Learning to Support Decisions with Sparse Data
Bill Eddy Verification of on-line forms for Census Bureau using record-matching and geographical data
Scott Fahlman

Symbolic learning from single or a few examples

Using symbolic knowledge to guide statistical learning (and vice versa)

Learning by reading

Christos Faloutsos Visualization and interaction with large graph
Matt Gormley Machine Learning for Natural Language Processing
Kotaro Hara & Jennifer Mankoff Using machine learning, computer vision, and crowdsourcing to evaluate daylight levels of buildings for designing urban technologies. Student must have a strong background in Machine Learning. Knowledge of Computer Vision (image segmentation and 3D geometry) and Web Development (crowdsourcing tool deployment) are useful.
Ken Koedinger Machine learning to model human learning and automate assessment and tutoring of math, science, and computer science in online courses

Jian Ma

Decoding what's encoded in the human genome using machine learning.

Jennifer Mankoff, Jeff Bigham & Scott Hudson

Using ML + Crowd to quality check data analysis procedures
Jack Mostow Fit and tune student models using performance data from an individual student or small population of students.  The context is the XPRIZE Global Learning Challenge to develop an open source Android app for African children to learn basic reading, writing, and numeracy without human teachers.
Robert Murphy
We work on learning statistical generative models for the structure and distribution of organelles within cells directly from microscope images. There is a review article available at that summarizes our past work on the CellOrganizer project. The fundamental machine learning question is to be able to decide which of various modeling methods provides a better representation of the information in images of populations of cells.
To enable screen reader support, press shortcut Ctrl+Alt+Z. To learn about keyboard shortcuts, press shortcut Ctrl+slash.
Jeff Schneider Recognizing bird species from audio of their singing
Aarti Singh

Online and active optimization for material discovery

Personalized prediction of cardiovascular risk from brain scans

Michael Tarr Predicting neural vision data using deep networks
Manuela Veloso

Learning patterns from logged data from a mobile service robot, CoBot