I gave a short talk at the 2021 National Homeland Security Conference. See https://www.nationalhomelandsecurity.org/. The event was held from August 30 through September 2.

The event Web site states, "The National Homeland Security Conference brings together professionals in Homeland Security, Law Enforcement, Fire and Emergency Management. They include officials in federal agencies, nonprofit agencies, business owners, universities and decision makers to learn about emerging trends in homeland security."

I estimate the event had about 2,000 attendees but I could be way off on that guess. Because of the way the conference was laid out, it was difficult to see how many people were there at any given time. Most of the attendees I talked to worked for State and Federal government agencies, but there were people from industry and academia too.

My talk was titled "AI and ML for Cyber Threat Prevention". I described different techniques for anomaly detection, such as deep neural autoencoder reconstruction error.

My talk was an out-of-band presentation and was set up by one of the sponsoring companies for a private audience during a break from the regular session times. The conference focused mostly on things like planning and logistics, rather than technical topics like mine, and so my talk wouldn't have had broad appeal. But the audience members who did show up were very enthusiastic and asked good questions.

One of the main reasons I speak at conferences is to gain insights into trends in the business and government communities. This information helps me prioritize the projects I do at work. My key takeaway from the 2021 NHSC event was that many government organizations seem to be mostly reactive -- that is, planning what to do when something bad happens -- rather than proactive -- trying to prevent disasters from happening in the first place.


Seconds before disaster.