Apple

I wear many hats including data storyteller, data scientist, data analyst, data modeler, data engineer, data architect, data wrangler, data detail manager, data miracle worker, data fire fighter, data bug squasher, all things data.

Health Data Scientist / Machine Learning Engineer

Fresenius Medical Care North America

•Trained and productionized multiple predictive models in Python with machine learning (scikit-learn/ XGBoost) to improve care of over 180,000 patients with kidney failure who are treated with dialysis
•Built, deployed, and maintained several custom Tableau dashboards to improve efficiency of clinical workflows based on predictive models
•Defined actionable metrics, performed analysis in Python, and communicated insights to C-suite, data scientists, data analysts, physicians, and nurses to improve these model-based workflows
•Acquired, cleaned, and integrated multiple EMR data sources (e.g. treatment vitals, monthly labs, and free-text clinical notes) using SQL and Python
•Mentored a junior data scientist and managed a data science intern
•Collaborated with data analysts and IT professionals to build a robust modularized data pipeline and to automatically score multiple predictive models daily or weekly
•Led hands-on NLP workshop at ODSC East 2019 and ODSC West 2019
•Conference abstracts: https://bit.ly/2p6atAI https://bit.ly/2NvRNUg
•Media article: https://bit.ly/3autf70
•Journal article: https://bit.ly/2JogYZq

Senior Data Scientist

• Implemented, piloted, and deployed an XGBoost model in Python to identify which patients on hemodialysis were at the highest risk of imminent hospitalization, and highlight each patient’s top risk factors
• Collaborated with data analysts, IT professionals, and nurses to automatically extract data, generate predictions, and display hospitalization risk in a daily Tableau dashboard to help care for 20,000 patients
• Presented on our machine learning efforts at 4 conferences, including 1 hands-on NLP workshop at ODSC West 2018

Data Scientist II

Northeastern University

• Taught 4 sessions of a 6 week course in Predictive Analytics for the Master of Professional Studies in Analytics. Through this course the students had the opportunity: 
   1) Investigate the basics of predictive analytics utilizing Python through real-world problems
   2) Employ a variety of predictive modeling techniques (classification, regression, clustering) and select the best approach for a real-world problem
   3) Create features suitable for predictive models
   4) Evaluate the performance of predictive models
   5) Iterate a model to adjust the predictive efficacy 
   6) Communicate the outcomes of a predictive analysis model that addresses a real-world business problem

Data Analytics Part-Time Faculty Member

Insight Data Science

• Participated in a summer data science fellowship program
• Consulted with a company to build a web app to automate genetics counseling by predicting the intended question from a patient's input question, identifying the health condition and providing a response
• Implemented in Python a multi-class logistic regression with scikit-learn using natural language processing (bag-of-words and tfidfvectorizer)
• Acquired, cleaned and combined data from multiple sources (company's data, Kaggle Quora competition, Genetics Home Reference) using Jupyter Notebook, NLTK, and Pandas.  Stored the data in multiple PostgresSQL databases
• Deployed the web app on Amazon AWS with Flask

Insight Health Data Science Fellow

Harvard University

• Led experiments to optimize a robotic exosuit for walking, reducing the metabolic cost by ~20%
• Provided an actionable recommendation that multiple metabolic measurements are needed in future experiments to reduce noise due to statistical analysis of within subject and between subject measurement variability
• Advised multiple graduate students on design and statistical analysis of their experiments
• Assisted writing 3 publications currently in review in high impact journals

Post-Doctoral Fellow

Johns Hopkins University

Thesis: Learning to place one foot in front of the other: investigation of action and perception in human split-belt walking. 

• Designed and conducted experiments with 400+ human subjects (healthy and patients with neurological diseases) to study how people learn to walk on a treadmill with two belts that move at different speeds
• Collaborated with two post-docs to develop a model to characterize what gait parameters are learned while adapting to split-belt walking 
• Developed a predictive model based on sensory cancellation to estimate the changes in perception of walking speed that results from motor learning
• Wrote MATLAB code to process and analyze 400+ GB of motion capture data acquired from Optotrak. 
• Analyzed data using repeated measures and mixed model ANOVAs in MATLAB
• Created multiple visual environments to provide feedback and assay perception using Vizard (Python based virtual reality software)
• Wrote 2 first author journal papers and assisted with 3 journal papers

Doctoral Student / Post-Doctoral Researcher

• Developed and simulated algorithms for error propagation and motion planning in stochastic robotic systems in MATLAB
• Wrote 2 first author conference papers and assisted with 1 journal paper

Visiting Graduate Student

Northwestern University

Thesis: The ParkourBot: modeling, experimental results and optimization.

• Programmed a two-legged robot with a PID controller to dynamically climb up and down
(for video see https://nxr.northwestern.edu/research/dynamic-locomotion/parkourbot)
• Created an optimization algorithm for motion planning of a dynamic climbing robot in MATLAB and Mathematica
• Constructed a custom printed circuit board and programmed the robot with PIC32
• Developed and experimentally validated a model of a planar biped  hopping robot
• Wrote 3 first author conference papers and assisted writing 1 journal paper

Graduate Student

NASA Jet Propulsion Laboratory

Engineered mass models and rapid prototypes with NX Unigraphics (CAD software) for the Curiosity Mars rover.

Cooperative Education Student

NASA Langley Research Center

Reduced data processing time with MATLAB for the Orion Crew Exploration airbag landing experiments

"Satisfaction of one's curiosity is one of the greatest sources of happiness in life" -Linus Pauling

I am a machine learning engineer who is passionate about using data science to build new products that improve healthcare. 

Taplio

Andrew Long, PhD's Linkedin Analytics

Andrew Long, PhD

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Andrew Long, PhD's Linkedin Analytics

Andrew Long, PhD

Want to drive more opportunities from LinkedIn?

Andrew Long, PhD's Linkedin Analytics