A panel discussion on artificial intelligence will highlight a luncheon scheduled for noon to 2 p.m. on Thursday, Jan. 24 at Saratoga Springs City Center. "The AI Opportunity: Developing an AI Ecosystem in Upstate New York" will tell why artificial intelligence matters, and what opportunities exist locally and regionally. Panelists will share ideas, experiences, and viewpoints about AI technology, research and development, ethics, and policies. There will be time to network with local leaders, industry experts, and community stakeholders following a discussion and question-and-answer session. This event is the first in a series of economic development "Lunch and Learns."

Four years ago, Alameda County, California's purchase of two drones for use by law enforcement was controversial. Now, the Alameda County Sheriff's Department has six drones, and their use is routine. So said Tom Madigan, a commander at the Alameda Sheriff's Office, to drone industry representatives and other law enforcement officials gathered at Drone World Expo in San Jose, Calif., last week. The Alameda County drone program has been fully operational for only about two years, Madigan said. In that time, he indicated, the Alameda Sheriff's Office has flown drones 700 times as part of 175 real-world missions, including search and rescue, fire scene surveillance, homicide scene analysis, and providing eyes in the sky during high-risk tactical operations.

We aim to produce predictive models that are not only accurate, but are also interpretable to human experts. Our models are decision lists, which consist of a series of if...then... statements (e.g., if high blood pressure, then stroke) that discretize a high-dimensional, multivariate feature space into a series of simple, readily interpretable decision statements. We introduce a generative model called Bayesian Rule Lists that yields a posterior distribution over possible decision lists. It employs a novel prior structure to encourage sparsity. Our experiments show that Bayesian Rule Lists has predictive accuracy on par with the current top algorithms for prediction in machine learning. Our method is motivated by recent developments in personalized medicine, and can be used to produce highly accurate and interpretable medical scoring systems. We demonstrate this by producing an alternative to the CHADS$_2$ score, actively used in clinical practice for estimating the risk of stroke in patients that have atrial fibrillation. Our model is as interpretable as CHADS$_2$, but more accurate.

Graphical Markov models use graphs, either undirected, directed, or mixed, to represent possible dependences among statistical variables. Applications of undirected graphs (UDGs) include models for spatial dependence and image analysis, while acyclic directed graphs (ADGs), which are especially convenient for statistical analysis, arise in such fields as genetics and psychometrics and as models for expert systems and Bayesian belief networks. Lauritzen, Wermuth and Frydenberg (LWF) introduced a Markov property for chain graphs, which are mixed graphs that can be used to represent simultaneously both causal and associative dependencies and which include both UDGs and ADGs as special cases. In this paper an alternative Markov property (AMP) for chain graphs is introduced, which in some ways is a more direct extension of the ADG Markov property than is the LWF property for chain graph.