That plays a large role when putting AI into practice. However, my understanding is "expert systems" rely solely on rule-based systems rather than machine learning.
That's the problem with the word "intelligence" being ill defined that he talks about on the video. An expert system can be categorized as an AI system, since it can solve problems and make decisions, as long as you give it the exact rules/principles to do so. But generally when people talk about AI they're referring to systems that have a model obtained through machine learning techniques on it's core, and that's what he's explaining: "AI & Machine Learning". One of the biggest appeals of machine learning methods is that given that you have enough data it can give you a model to power a system which can be used instead of an Expert System, but without needing experts working for long periods to code it line by line.
If you want me to give a talk at your company, go here: www.shawhintalebi.com/ai-workshops
Thanks Shaw. This was very useful. I shared it with some of my friends too.
Thanks for sharing :) Glad it helped!
Defining 'intelligence' and the similarities between 'human mental models' & 'ML models' hits it home.
So broken down and easy to understand! Thank you Shaw, super helpful for AI-dummies like me!
Glad it was clear and helpful!
Nice explainer! Thanks!
Glad you liked it :)
Hi Shaw, a very crisp and clear video.
Thanks Sanjay! Glad it was clear :)
Can you do a course on Markov / Semi Markov / Hidden Markov / Semi Hidden Markov models please.
Great suggestions! Added it to my list :)
@@ShawhinTalebi Great. Many thanks for this.
I rather use the proper term "Expert System."
That plays a large role when putting AI into practice. However, my understanding is "expert systems" rely solely on rule-based systems rather than machine learning.
That's the problem with the word "intelligence" being ill defined that he talks about on the video. An expert system can be categorized as an AI system, since it can solve problems and make decisions, as long as you give it the exact rules/principles to do so. But generally when people talk about AI they're referring to systems that have a model obtained through machine learning techniques on it's core, and that's what he's explaining: "AI & Machine Learning".
One of the biggest appeals of machine learning methods is that given that you have enough data it can give you a model to power a system which can be used instead of an Expert System, but without needing experts working for long periods to code it line by line.
Please say in hindi.
I would if I could 😅