valentino.io

Data Scientist based in Los Angeles

About me.

Hi! Thanks for reading. My name's Valentino. I'm a Data Scientist at NASA's Jet Propulsion Laboratory (JPL), using machine learning, web development, and effective communication to improve the laboratory's operational efficiency. My work currently revolves around tool building, unsupervised learning, and graph networks. I aim to improve the effectiveness in which JPL conducts missions focused on Earth or Solar System exploration in my own small way.

I also do consulting work for firms interested in using data science and machine learning for their products, services, or to guide strategy. Feel free to reach out.

When I'm not at JPL or doing consulting work, I enjoy hiking, working on research and contributing to open-source projects, cooking (but not always), or volunteering. Gardening, hiking and working with Raspberry Pis and Arduinos, and spending time with my dog are fun pastimes, too.

More in-depth info can be found on my resume if that's what you're looking for.

Explore my work.

MIEDA

data science: graph algorithms

Merging of set-containing Intervals Efficiently with a Directed-graph Algorithm.

fork the repo

Spacecraft Anomaly Detection

data science: time-series anomaly detection

Detection of anomalies in spacecraft telemetry using deep learning.

fork the repo

Henosis

data science: machine learning, recommender systems

A generalized form field recommendation system for categorical machine learning models.

fork the repo

PyNomaly

data science: outlier detection, anomaly detection

A Python implementation of Local Outlier Probabilities (LoOP) introduced by Kriegel, Kröger, Schubert, and Zimek.

fork the repo

72.100

speculative design

72.100 is a speculation on a reality we may find ourselves in, should we ignore the important and vital issue of water quality.

learn more

Chicago Sunrise Data

data science: acquisition

Colors and brightness of the Chicago sunrise as data from five different locations.

fork the repo

Hey Waldo

data science: acquisition, deep learning

Labeled images of the Where's Waldo puzzle for use in classification and image recognition problems. Color, grayscale, and black or white images are included.

get the data

A Data Driven Marathon

data science: acquisition, visualization

The city of Chicago is forward-thinking in regards to open access to civic data. We illustrate that marathon organizers and the city could improve the safety of the Chicago Marathon by analyzing data from health trackers.

read my article

kNN Imputation

algorithms

A naive implementation of the K nearest neighbors algorithm in R, used here for imputing values in missing data.

fork the repo

runstest-R

data science: R package

An implementation of the runs test (Wald-Wolfowitz test) in R. Tests whether autocorrelation is present in the residuals, which is useful for time series models.

fork the repo

JPL Facilities Search

data science: acquisition, visualization

Now open source, this tool allows for intuitive understanding of space constraints and potential opportunities for personnel moves that improve organization efficiency.

fork the repo

HERE Maps

data science: machine learning, visualization

Developed visualization tools and machine learning models for determining the spacial extent and criticality of roadway congestion from construction activity.

visit HERE site

dateCleanse.py

data science: cleansing

A Python-based tool that reformats 14 date formats into a single format for later use.

fork the repo

Julia Numerical Algorithms

algorithms

A collection of numerical algorithms written in the Julia language.

fork the repo

GE Transportation

data science: cleansing, machine learning, visualization

We used anomaly detection methods and data visualization to easily identify locomotives with abnormal maintenance behavior for the BNSF fleet.

read the report

Field Horizon

web design

A representation of Chicago sunrise data, collected and visualized between students from the School of the Art Institute of Chicago and Northwestern University.

visit site

Experience

NASA Jet Propulsion Laboratory: Pasadena, CA

Data Scientist III, March 2020 - present

- Technical leader responsible for guiding research and development of data, analytics, and machine learning products.
- Continued research and development of anomaly detection approaches and applications of graphs (e.g. directed acyclic graphs) in support of internal efforts and work in support of the DARPA SafeDocs program.

Principal Investigator, September 2018 - present

- Responsible for overall technical strategy for multi-year fault detection and alarm management effort that utilized numerous machine and deep learning approaches, from Principal Components Analysis (PCA) to various time-series deep learning architectures.
- Responsible for managing and mentoring co-investigators and interns, as well as project budget and communication with external and internal stakeholders.

Data Scientist II, March 2018 - March 2020

- Research and development of anomaly detection systems for the Soil Moisture Active Passive (SMAP) spacecraft using machine and deep learning approaches for time-series applications such as LSTMs.
- Developed numerous full-stack prototype internal product or services on top of Natural Language Processing (NLP) and classification approaches, such as classification-based recommendation systems and tools to assess water resource variable (e.g. drought management, water reservoir height) priorities across water resource agencies. Responsible for architecture design, model training, API development, and front-end interfaces of numerous product or service prototypes.

Data Scientist, February 2017 - February 2018

- Lead the development effort from early concept to near-production classification-based recommendation system for the Problem Reporting System (PRS).
- Developed small scale prototype of a tool that utilized graph theory to assess employees’ relationship strength’s with external companies and organizations and released the Python open-source software PyNomaly.

IT Data Science Intern, June 2016 - August 2016

- Developed an application that aids facilities planners and other organizations throughout the facility.
- Worked on internal data science consulting projects for various organizations across lab.

Alcoa (Materials Management, Global Primary Products): Knoxville, TN

Intern, February 2013 - April 2014

- Performed market analysis of primary aluminum market as a member of the strategic marketing team.
- Developed a predictive model for US & Canadian extruded products directly affecting 2014 strategy.

Details regarding additional experience may be viewed on the PDF resume, linked below.

Publications

Papers

Building a Wide Reach Corpus for Secure Parser Development

Allison T., Burke W., Constantinou V., Goh E., Mattmann C., Mensikova A., Southam P., Stonebraker R., Timmaraju V.

41st IEEE Symposium on Security and Privacy Workshops - LangSec Workshop. 2020.

An Overview of Chicago Marathon Prehospital Care Demographics, Patient Care Operations and Injury Patterns

Chan J., Constantinou V., Fokas J., Van Deusen Phillips S., Chiampas G.

Prehospital and Disaster Medicine. Cambridge University Press. 2019.

PyNomaly: Anomaly detection using Local Outlier Probabilities (LoOP)

Constantinou V.

Journal of Open Source Software, 3(30), 845. 2018.

Detecting Spacecraft Anomalies Using LSTMs and Nonparametric Dynamic Thresholding

Hundman K., Constantinou V., Colwell I., Laporte C., Soderstrom T.

Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2018.

Applicability of Consumer Activity Monitor Data in Marathon Events: An Exploratory Study

Constantinou V., Felber A., Chan J.

Journal of Medical Engineering & Technology. 2017.

Talks

Data Science at NASA's Jet Propulsion Laboratory

Constantinou V. Online via Zoom for Northwestern University students and alumni. May, 2020.

Anomaly Detection for Spacecraft/Rover Telemetry Using Long Short-Term Memory Network

Constantinou V, Colwell C. Ground Systems Architectures Workshop. Renaissance Los Angeles Airport Hotel. March, 2020. El Segundo, CA.

Telemanom: an Extensible Framework for Time-series Anomaly Detection

Constantinou V. NSET Geoinformatics Engineering Group Seminar. Oak Ridge National Laboratory. March, 2019. Oak Ridge, TN.

Henosis - a generalizable, cloud-native Python form recommender framework for data scientists

Constantinou V. Workshop at the the Open Data Science Conference - East. May, 2018. Boston, Massachusetts.

Pre-Hospital Care at a Mass Endurance Event: The Chicago Marathon Experience

Goyal A., Constantinou V., Fokas J., Van Duesen Phillips S., Chan J., Chiampas G. Oral presentation at the WADEM Congress on Disaster and Emergency Medicine. April, 2017. Toronto, Canada.

Posters

Enabling Earth Science Research Through Use of Aura MLS Data with Outside Science Data Products

Knosp B., Constantinou V., Livesey N. Poster at the the 2019 Aura Science Team Meeting. 2019. Pasadena, CA.

Pairing NASA Capabilities and Western US Water Resource Stakeholders with Data Mining and Machine Learning

Lee C., Constantinou V., Mishra A. Poster at the the American Geophysical Union (AGU) Fall Meeting. December, 2018. Washington, DC.

Detection and Interactive Exploration of Telemetry Anomalies

Constantinou V., Laporte C., Colwell I., Hundman K. Poster at the the American Geophysical Union (AGU) Fall Meeting. December, 2018. Washington, DC.

Education

M.S., Analytics: Northwestern University, McCormick School of Engineering

- Exploratory and visual analysis of medical data from past Chicago Marathons, providing critical information to Chicago Event Management for use in medical staffing. Experimental design and analysis of consumer activity monitor data in order to understand and evaluate potential use cases toward mass running events.
- Contributing team member of a nine month practicum project with GE Transportation, involving both unsupervised and supervised machine learning methods, in addition to visualization in Tableau. Designed and evaluated predictive models using random forests, gradient boosted trees, neural networks, among others.
- Ten week capstone design project with HERE Maps. Developed visualization tools and machine learning models for determining the spacial extent and criticality of roadway congestion from construction activity for the German Bundesauthobahn 5.

B.S., Economics: The University of Tennessee - Knoxville, Haslam College of Business

- Studies abroad at the University of Melbourne and the London School of Economics.
- Alan & Wendy Wilson, CIE, David Smith Chambers, and HOPE Scholarship recipient.

Knowledge Areas

- Anomaly Detection	- Data Visualization	- Dev Ops	- Natural Language Processing
- Data Mining	- Deep Learning	- Machine Learning	- Time-Series Modeling

Languages and Frameworks

- Angular	- Elasticsearch	- Go	- R	- SQL
- CSS3	- JavaScript (vanilla or ES6)	- LaTeX	- React / JSX	- Tika
- d3.js	- Julia	- Python	- Spark	- Torch

Volunteering

Office of 27th District of California - Congresswoman Judy Chu

Judge - Congressional App Challenge: November 2019

Institute of Electrical and Electronics Engineers (IEEE)

Reviewer - IEEE Intelligent Systems: May 2020

Reviewer - 2020 IEEE Aerospace Conference - Software and Computing Track: November 2019

Journal of Open Source Software (JOSS)

Reviewer - OrNet: a Python Toolkit to Model the Diffuse Structure of Organelles as Social Networks: December 2019
Reviewer - Graph Transliterator: a graph-based transliteration tool: September 2019
Reviewer - webweb: a tool for creating lightweight network visualizations in the browser: May 2019
Reviewer - Netwulf: Interactive visualization of networks in Python: May 2019
Reviewer - Implementation of the Robust Random Cut Forest algorithm for anomaly detection on streams: March 2019

General Assembly

A Day in the Life of a Data Scientist (Downtown Los Angeles): March 2019

STEAM:CODERS

Family Code Night (ages 6-12): October 2018

Introduction to JavaScript (ages 12-16): October 2017 - November 2017

Arduino Robotics (ages 8-11): April 2017 - May 2017

Let's create together.

I'm always looking for the next challenging and interesting project to work on, whether it be in machine and deep learning, data visualization, web design, or another creative endeavour. Let's create together, solve a problem, and have some fun.

Dive in.

Detecting Anomalies in Spacecraft Telemetry

Using Long Short-Term Memory (LSTM) models.

NASA JPL Magnetosphere Project

Clustering predicted states of the Earth's magnetosphere.

A Visual Look at Under and Overfitting using U.S. States

Exploring concepts through decision surfaces.

Intelligently Assisted Form Fields with Henosis

A generalized categorical form field recommender system.

Why Context Matters: Lessons from the Chicago Marathon

Context is key toward understanding data and applying appropriate analytical methodologies.

A Data Driven Marathon