Harrison Kane’s Data Science Portfolio
AWS Cloud Deployment
Store data and deploy predictive models to AWS Cloud for security, monitoring, and a User-Friendly UI
Custom ML Software Development
Build an automated ML tool to predict a business outcome based on a set of input features.
Data Warehousing & Engineering For Business Intelligence & Data Science
Wrangle, clean, and process myriad data sources into a Machine Learning-friendly format with SQL and Python
Robust Cross-Industry Experience
4 Years Total as a Consultant and In-House Team Member
Fluent in Spanish, Proficient in Chinese
University of Michigan Princeton in Beijing
Technology to Business Liason
Translator of Statistical Findings into Business Narrative
4 Key Steps in Machine Learning
A high-level distillation of the common processes followed in an applied data science context.
The Process of Data Science
The breadth of Data Science and scope of its applications in business, heathcare, and beyond makes it difficult to distill into fundamental steps ubiquitous to all data science use cases. However, here are 4 steps that I have found to be mostly consistent both in my work as a data science consultant as well as an in-house position.
Data Collection and Pre-Processing
Collect and store relevant historical data tables in a data warehouse. Join, aggregrate, and correct data to create a modeling dataset for analysis.
Exploratory Data Analysis
Generate visualizations to highight KPIs relevant to the client’s business performance.
Model Development & Iteration
A series of statistical models are trained on the historical dataset, then compared for accuracy to find the best model.
Model Deployment, Integration, and Maintenance
The best-fitting model is deployed to the cloud and invoked with new data points via API to make predictions.
