Automating Data Science – Insights and Best Practices

Feature Engineering is what separates good machine learning outcomes from great machine learning outcomes.  Part art and part science, feature engineering combines domain expertise with data science acumen to transform raw data into the variables that drive machine learning models.

In this second webinar in our six-part series, Socure’s VP Data Science, Pablo Abreu, discusses the methods used in feature engineering to go from raw data (discussed here Data Exploration - Improving the Foundation) to highly predictive features. More specifically, Pablo will share the approach Socure takes to build proprietary predictors to boost machine learning model performance.

In this webinar, Pablo will share:

• General concepts and approaches to feature engineering
• Typical challenges encountered with feature engineering
• Several best practice approaches to consider to maximize success
• Considerations for manual vs. automated approaches to feature engineering

In this first of six webinars, Pablo starts at the beginning and discusses different methods for data exploration. As the goal of data science is to extract knowledge and insights from data, it makes sense that we start with data and explain the process used for the initial analysis of data.

In this on-demand webinar, Pablo shares:

• General concepts and approaches to data exploration
• Typical challenges encountered with data exploration
• Several best practice approaches to consider to maximize success