Throughout the data e cosystem, organizations are beginning to realize the worth of an enterprise information fabric that uses semantic technology for business understandin g to provide uniform access to all data assets, structured and unstructured , regardless of how dispersed they are.
\nOne of the innate benefits of using a semantic approach to construct a data fabric is the ability to h armonize all data into an Enterprise Knowledge Graph, predicated on busines s meaning, that’s perfect for machine learning feature engineering. T raditionally, data preparation was a data science and machine learning bott leneck; it was so time-consuming it limited the impact of this valuable tec hnology. This talk will discuss how knowledge graphs can be used to acceler ate this process in four key ways to maximize machine learning engineering results:
\n\n Training Data: graph algorithms provide a far richer source of training and operational data than those of other analytic approaches.\n Unstructured Data: graphs are difficult to surpass for harmonizing unstruct ured and structured data to maximize the potential for predictive signal fr om more sources.\n Feature Engineering: semantic graphs drastically decrease the time and effo rt of feature engineering, partly due to automated query and transformation generation, leading to automated feature synthesis.\n Provenance: When operationalizing machine learning models, graphs provide a n immutable data lineage chain to retrace data’s journey from its ini tial model testing phase making it easier to recreate that journey when ML models are put into production.\n\n DTSTART:20181010T121000 SUMMARY:Organizing and Accelerating Enterprise Machine Learning Projects wi th Enterprise Knowledge Graphs DTEND:20181010T123959 LOCATION: See Description END:VEVENT END:VCALENDAR