Anzu (あんず /a.N.zɯ/)

Software/Front-end Developer Visual Analytics & HCI Researcher Amateur Bathroom Singer

About Me

I'm a current Master's student studying Computer Science at Tufts Unviersity, graduating in May 2017. I previously obtained a Bachelor's degree from Tufts University with a double major in Computer Science and Psychology. I have strong communication skills in both English and Japanese and am able to utilize knowledge from living in diverse cultures in order to connect with others and develop new and eclectic ideas :) I'm committed to quality and excellence and am eager to learn new skills. I'm interested in working in IT fields as well as positions that applies creativity and visual arts designing.




IEEE Transactions on Visualization and Computer Graphics

In this paper, we report the iterative design process and the findings of developing the Prognosis Assessment for Conservative Treatment (PROACT) tool, a personalized health risk communication tool for localized prostate cancer patients. PROACT utilizes two published clinical prediction models (CPMs) to communicate the patients’ personalized risk information of prostate cancer and compare treatment options. In collaboration with the Maine Medical Center, we conducted two rounds of evaluations with prostate cancer survivors and oncologists to identify the design elements and narrative structures that would increase the users’ comprehension of their health information.

Our results indicate that visualization can be an effective means to communicate complex risk information to patients with low numeracy and visual literacy. However, the visualizations need to be carefully chosen to balance readability with ease of comprehension. In addition, due to patients’ charged emotional state, an intuitivenarrative structure that considers the patients’ information need is critical to aid the patients’ comprehension of their risk information.

EMR Security Vulnerability

COMP 116 Security Final Project

As more healthcare institutions utilize electronic medical record software (EMR) to store patient records, the number of healthcare data breaches is also on the rise. In addition to the obvious threat to privacy, there are malicious ramifications to medical information cyber attacks, with medical records valued higher on the black market than credit card or Social Security Numbers (SSN). The stolen records can be used for spamming, identify fraud, prescription and service theft, billing fraud, and in its worst case, lead to death. While electronic medical record software companies need to improve their system security, the users of the software are also at fault as this vulnerability is exacerbated by the lack of training and awareness from both the hospital employees and patients. This paper will explore the causes and ramifications of healthcare information breaches as well as suggest some possible defense tactics to prevent medical information cyber attacks.

Comparing Supervised Machine Learning Algorithms for Emotion Recognition Performance

COMP 135 ML Final Project

I compared 8 machine learning algorithms for human emotion recognition using Scikit-learn's python library. The algorithms were Linear Discriminant Analysis, Support Vector Machine, Stochastic Gradient Descent, K-Nearest Neighbors, Naive Bayes, Decision Tree, AdaBoost, and Label propagation. I used the Extended Cohn-Kanade Dataset (CK+), which is a database developed by Carnegie Mellon Univer- sity and the University of Pittsburgh, and it contains facial expression data that can be used to study automatic facial expression detection. Specifically, I used the Active Appearance Models Landmarks (AAMs), Facial Action Coding System (FACS) Action Unit (AU) Labels, and the emotion labels for the video sequence.



Deferred shader

Deferred shading is a form of scene rendering: all of the components that contribute to shading (e.g., normals, position, color, depth) is rendered onto a texture first (G-buffer) and then the shading is calculated per pixel using the G-buffer textures. Unlike ray casting, which could have long rendering times, deferred shading allows for not just rendering of many lighting but also rendering of moving lights at a faster speed.

healthy habits

Healthy Habits

Personal health and fitness covers a wide variety of areas. There is physical fitness and the much neglected mental health. How can you create an app to address both of these areas and inspire people to strive to be fit in both areas?

microburst game

Lyrical Pie Bubble Visualization

An exploratory visualization demo that relates word frequencies across genres. The diameter of the bubbles corresponds to the frequency of a specific word relative to the total word occurrences, and the pie arc lengths correspond to frequencies in their respective genres.

microburst game


Ever felt like you were about to implode after a hefty mexican meal? Is there an internal war going on your stomach? Is there a growing microbe feeding on what you eat? Microburst is the game for you!

3D Design

haunted mansion

Haunted Mansion

Modeled using Blender, this scene is a haunted mansion that includes objects with materials, textures, and a fly-around of the structure.



Modeled using Blender, this scene is a wind sculpture of fairies that includes objects with materials textures, and a fly-around of the sculpture.



Modeled using Blender, this scene is a wind sculpture of fairies that includes objects with materials textures, and a fly-around of the sculpture.

Recent Updates



Picture of Anzu