This is simply a running list of the things I work on outside of school and work.
Machine Learning Projects
- Neurofinder Calcium Imaging Segmentation Challenge
- ISBI 2012 Electron Microscopy Segmentation Challenge
- Reservoir Computing Term Paper
Web Development Projects
- Meetup.com RSVP Visualization
- Dynamic Adaptive Neural Network Array Visualization
- Animated Finite State Machine Converter
- IEEE IOT Hackathon
- Maritime Piracy Mapping and Analysis
- Freelance Websites
Neurofinder Calcium Imaging Segmentation Challenge
ISBI 2012 Electron Microscopy Segmentation Challenge
- Fully-convolutional segmentation based on UNet.
- End-to-end implementation places top 30 on leaderboard (search "ORNL_Klibisz").
- Github Repository
Reservoir Computing Term Paper
- 14-page term-paper introducing Reservoir Computing, Liquid State Machines, and Echo State Machines.
- Paper, slides.
- Tensorflow classifier pre-trained on ImageNet classifies 1-second frames from arbitrary Youtube videos.
- Web applicatino lets the user jump to frames where specific labels occur.
- Built at VolHacks 2016 hackathon
- Github Repository, TensorTube.tech
Meetup.com RSVP Visualization
- Real-time visualization of streaming data from the Meetup.com open events RSVP API.
- Visualize events as a map with markers and bar charts showing the most frequent countries, states (USA), and names.
- Source code: https://github.com/alexklibisz/meetup-viz
- Talk description: http://www.meetup.com/Knoxvillejs/events/229148319/
Co-founder, september 2015 - March 2016
- StudyLoop is a web and mobile application that gives students an easier way to communicate with one another. We use the specific course schedule for a university to make it possible for students to join their courses within the app and avoid the hassle of setting up group messaging or other collaboration. Conversations are structured as loops, simply individual message groups, pertaining to specific assignments or tasks. This helps promote organized, productive conversation.
- For StudyLoop, I built the web application using React.js and all of the data services (push notifications, email notifications, complex data logic, email verification, image and file attachments) using Firebase, AWS Elastic Beanstalk, AWS SNS, AWS SES, AWS S3, and AWS Lambda.
- I worked with co-founder Sam Rose and developer Chris Martin. They focused primarily on the iOS app.
- The app launched in February 2016 for web and the iOS store and has approx. 160 users as of March 2016.
- To demo the app you can go to http://studyloop-stage.firebaseapp.com and login with email firstname.lastname@example.org and password test123.
Press and Links
- StudyLoop on thei iOS app store
- Sam Rose pitching StudyLoop at Knoxville Entrepreneur Center Demo Day, December 2015
- CodeWorks Demo Day targeted entrepreneurs - Knoxville News Sentinel
- CodeWorks Team 1: StudyDraft - Teknovation.biz
Dynamic Adaptive Neural Network Array Visualization
August 2015 - December 2015
- The Dynamic Adapative Neural Network Array is a neuromorphic network hardware implentation created by the Neuromorphic Computing research group at the University of Tennessee.
- DANNA-View is a visualization web-app that I built for visualizing the DANNA networks as part of an independent study with Dr. James Plank in the fall 2015 semester.
- The tool uses definition files for a DANNA to produce interactive visualizations and heat maps for network activity.
- The visualizations are used for better understanding the networks, rendering graphics for publications, and high-level debugging of the networks.
Animated Finite State Machine Converter
January - April 2015
- Visual Finite State Machine Converter built as an honors project for COSC312, Spring 2015.
- Allows the user to interactively build a Nondeterministic Finite Automaton and visually convert it to an equivalent Deterministic Finite Automaton.
- Used D3.js and Angular.js to create the visualization and controls.
IEEE IOT Hackathon
April 2015, San Jose, CA
- On April 26, 2015 I participated in the Cisco/WISE IEEE Hackathon in San Jose, CA.
- After forming teams, we were given the opportunity to build a useful IoT device using a series of hardware sensors provided by the company Phidgets.
- My group prototyped a prescription pill bottle sensor system that keeps track of how many pills are in the bottle and when the bottle was last opened and makes this information available with real-time updates online.
- We won the “Best Value to Business” award and a $1000 prize.
Maritime Piracy Mapping and Analysis
August 2014 - August 2015
- Web app used to filter, map, and graph ~7000 maritime piracy incidents dating back to 1993.
- Working under Dr. Brandon Prins, UTK, as part of Political Reach, State Fragility, and the Incidents of Maritime Piracy and Pirate Organizations, 1995-2016 research initiative.
- Independent programmer on this project, responsible for full-stack development.