Daniel Hartley
Data Analyst/Data Scientist
H
Data Analysis / Forecasting & Modeling / Agile / Python / Cloud / Process Design / Spark / Project Management / Process Improvement / Data Trends / SQL / KNN / Algorithms
Recent graduate of the University of Colorado and Galvanize with experience in data and trend analysis, risk identification, and data modeling to help companies implement process improvements and efficiencies. Highly-analytical and out-of-box thinker who thrives in high-pressure and complicated situations. Seeking opportunities to leverage these skills in a collaborative environment. I’ve developed an understanding of the following areas:
- Statistical Inference
- Regression
- Graph Theory
- Recommendation Systems
- Big Data Processing
EDUCATION
Bachelor of Arts, University of Colorado Boulder; Data Science Certificate, Galvanize - Denver; Neural Networks for Machine Learning, University of Toronto on Coursera
RELEVANT DATA SCIENCE EXPERIENCE
Identified areas of likely churn, enabling company to bolster customer usage. A ride-sharing company needed to ID users likely to drop their service. Employed EDA (exploratory data analysis) techniques to pinpoint trends and find which users were most likely to churn.
Found areas of likely fraud. Tasked to detect fraudulent transactions so future safeguards could be put in place. Developed model using previously recorded data to correctly ID new fraudulent transactions, extracted and analyzed text using a combination of term-frequency inverse document frequency model and Guassian Naïve Bayes model. Generated fraud prediction and added to feature matrix. Ran through Random Forest Classifier to make final predictions of low, medium, and high-risk transactions.
Improved previously calculated root mean square error (RMSE). From database of thousands of movies, users and ratings, tasked to create recommendation system. Using combination of Python and Spark, developed base model for the recommender system that was trained on subset of the data. Spun up node on AWS to test model on complete data set. Improved upon previously recorded RMSE.
Created model to identify commonalities in data. Incorporating Natural Language Processing, used tf-idf to vectorize each subset of data and NMF to split corpus of data into “topics”. Calculated cosine similarities in the database. Produced model that successfully identifies related data sets for analysis and reference for future cases.
LEADERSHIP ACTIVITIES
National Outdoor Leadership School (NOLS) Graduate: Backpacked through the Wind River Mountain Range over 30 days. Part of extraction team for injured hiker.
Founder and Officer: Dr. Who Club, University of Colorado.
Competitive Swimmer: Competed on seasonal and year-round swim teams.
ADDITIONAL EXPERIENCE
Display Technician, Tri-State Fireworks Inc., 2012 to Present. Set up and execute fireworks displays. Collaborate to deliver consecutive displays safely and on-time. Train new employees in processes and safety standards.
Recent graduate of the University of Colorado and Galvanize with experience in data and trend analysis, risk identification, and data modeling to help companies implement process improvements and efficiencies. Highly-analytical and out-of-box thinker who thrives in high-pressure and complicated situations. Seeking opportunities to leverage these skills in a collaborative environment. I’ve developed an understanding of the following areas:
- Statistical Inference
- Regression
- Graph Theory
- Recommendation Systems
- Big Data Processing
EDUCATION
Bachelor of Arts, University of Colorado Boulder; Data Science Certificate, Galvanize - Denver; Neural Networks for Machine Learning, University of Toronto on Coursera
RELEVANT DATA SCIENCE EXPERIENCE
Identified areas of likely churn, enabling company to bolster customer usage. A ride-sharing company needed to ID users likely to drop their service. Employed EDA (exploratory data analysis) techniques to pinpoint trends and find which users were most likely to churn.
Found areas of likely fraud. Tasked to detect fraudulent transactions so future safeguards could be put in place. Developed model using previously recorded data to correctly ID new fraudulent transactions, extracted and analyzed text using a combination of term-frequency inverse document frequency model and Guassian Naïve Bayes model. Generated fraud prediction and added to feature matrix. Ran through Random Forest Classifier to make final predictions of low, medium, and high-risk transactions.
Improved previously calculated root mean square error (RMSE). From database of thousands of movies, users and ratings, tasked to create recommendation system. Using combination of Python and Spark, developed base model for the recommender system that was trained on subset of the data. Spun up node on AWS to test model on complete data set. Improved upon previously recorded RMSE.
Created model to identify commonalities in data. Incorporating Natural Language Processing, used tf-idf to vectorize each subset of data and NMF to split corpus of data into “topics”. Calculated cosine similarities in the database. Produced model that successfully identifies related data sets for analysis and reference for future cases.
LEADERSHIP ACTIVITIES
National Outdoor Leadership School (NOLS) Graduate: Backpacked through the Wind River Mountain Range over 30 days. Part of extraction team for injured hiker.
Founder and Officer: Dr. Who Club, University of Colorado.
Competitive Swimmer: Competed on seasonal and year-round swim teams.
ADDITIONAL EXPERIENCE
Display Technician, Tri-State Fireworks Inc., 2012 to Present. Set up and execute fireworks displays. Collaborate to deliver consecutive displays safely and on-time. Train new employees in processes and safety standards.
Centennial, CO
Contact me for
Files
-
HartleyD_Resume_Final-.docx
vnd.openxmlformats-o, 20 KB