To earn a certificate of completion, the recipient spent approximately 20 hours in the following educational activities:
Attended live sessions and office hours.
Hands-on experience with Polygon Research apps, interacting with originations and servicing loan-level data, statistical records for people and households, and lender-specific insights. This was a guided experience through multiple use cases.
Exploration and choice of a follow-up a career-building data science learning path from among five choices: management, data engineering, coding, machine learning, and statistics.
The recipient completed the requirements for this intensive 7-week course featuring live sessions with office hours, weekly knowledge checks, hands-on experience with analytical tools, and choice of tailored learning paths.
The recipient of this certificate has developed a foundational framework that leverages the interconnected nature of AI / ML models, data, and housing finance domain expertise.
Specific activities and skills included:
Mastery in descriptive data science, working with 37 million rows of HMDA LAR data to filter, measure, and interpret geographical, loan production, and fair lending data.
Avanced analysis of U.S. Census 1-Year ACS microdata, encompassing 330 million people and 130 million households, to extract and reason about demographic and homeownership insights.
Choosing a learning path to explore and develop career path optimization.
The ability to interpret and analyze Agency MBS current mortgage market trends both year-to-date and over the trailing 12 months, and to distill findings into actionable insights for strategic decision-making in the housing finance industry.
Proficiency in filtering and drilling down into loan-level data to uncover granular insights about agency loans such as First Time Home Buyer segments.
Skill in identifying and understanding key trends, patterns, and anomalies within the Agency MBS market.
The recipient had hands-on experience in performing analysis using loan-level data and person/household-level statistical records in the following interactive SaaS apps:
1-HMDAVision
2-CensusVision
3-FHAPivot
In addition, the recipient had hands-on experience working with loan level data via:
-Snowflake
-SQL commands