NOW in 60: XAI series - Large Experience Models
XAI Series: Large Experience Model in 60 Seconds
In this short video, learn how Juniper's Large Experience Model (LEM) leverages billions of data points to predict performance issues on applications like Zoom or Teams almost instantly. It uses SHAP (Shapley Additive Explanations) to provide detailed visual explanations of AI model predictions, enhancing transparency and trust. Overall, it empowers IT teams to better understand and manage network operations.
You’ll learn
How Juniper's LEM improves AI-Native Networking
The role of SHAP in providing transparency into AI model predictions
How LEM can instantly identify and resolve app performance issues
Who is this for?
Host
Transcript
0:00 [Music]
0:06 at Juniper we're continually driving
0:08 improvements in AI native networking
0:10 with our large experience model or LM by
0:13 training on billions of data points our
0:15 LM can instantly predict performance on
0:17 apps such as Zoom or teams identifying
0:20 issues in seconds that would take months
0:22 to diagnose manually a key component of
0:25 our LM is shap or shly additive
0:28 explanations sha provides detailed
0:31 visual explanations of AI model
0:33 predictions assigning values to Features
0:35 based on importance this transparency
0:38 into the AI model helps operators better
0:41 understand and Trust AI outputs for
0:43 example shaft can instantly pinpoint a
0:46 misfigured VPN Gateway that's causing
0:48 latency issues and video calls by
0:51 combining network data with application
0:53 data AI models become more interpretable
0:55 and reliable this effectively
0:57 democratizes access to complex
0:59 operations enabling it teams to better
1:01 learn understand and drive change and
1:04 that's the building of our large
1:06 experience model in 60 Seconds