Intel’s Sean McPherson: AI Model Training Requires Human Input

Sean McPherson
Sean McPherson

Sean McPherson, deep learning scientist at Intel, wrote in an opinion piece published Monday on Federal Times that there are three strategies organizations can use to leverage the capabilities of artificial intelligence and one is encouraging humans to take part in AI model training.

“By inserting yourself into the continuous feedback process, you’ll have the opportunity to refine your trained models, evaluate results and data, and watch the model learn,” McPherson wrote. “You’ll get more comfortable with the AI process, and more confident in the technology as you watch it become more accurate and dependable.”

Another strategy is finding opportunities where AI can help build up the team. McPherson cited how AI systems could help organizations review untapped data that could prove useful in providing actionable insights and improving mission outcomes.

He said the third strategy calls for organizations to demonstrate the capabilities and inner workings of their AI models. He mentioned the Defense Advanced Research Projects Agency’s Explainable Artificial Intelligence project and Bayesian Deep Learning as some of the initiatives that seek to provide “clearer insights into how AI comes to certain conclusions.”

“With this information in hand, you’ll be able to decipher the decisions made by the AI algorithm and trace mistakes or inaccurate recommendations back to their root causes, and develop remediations through model adjustments or improved training data,” McPherson added.

You may also be interested in...

Seth Cutler CISO NetApp

NetApp’s Seth Cutler: AI Can Support Agency Data Security, Compliance Efforts

Seth Cutler, chief information security officer of NetApp, has said implementing artificial intelligence tools can help the public sector manage data protection, compliance and risks as agencies increase the pace of digital transformation projects to address cybersecurity threats.

Cybersecurity

DIU Taps CounterCraft to Help Deploy, Mature Threat Intell Prototype

The Defense Innovation Unit has awarded threat intelligence technology maker CounterCraft an other transaction agreement to mature the company’s counterintelligence tool in an effort to improve the protection of Department of Defense systems.

Palo Alto Networks

Palo Alto Networks Gets FedRAMP Approval for Three Data Security Products

Palo Alto Networks has added three cloud products to its suite of government-approved offerings under the Federal Risk and Authorization Management Program. The Palo Alto Networks Government Cloud Services now includes the Prisma Cloud, Cortex XDR and Cortex Data Lake platforms that have received moderate FedRAMP authorization with the help of the Federal Housing Finance Agency, the company said Monday.