How does AI learn from data

Nikhil Sharma
5 replies

Replies

Maxime Dolores
It depends on the type of task you’re trying to solve (supervised vs unsupervised) etc For supervised cases the AI predicts things compares with the expected result and uses the amount of error to enhance itself then tries again iteratively For unsupervised it’s a bit more tricky but there will usually be some way to measure progress or to converge towards a result by iterating as well
Nikhil Sharma
@m_dolr Thats interesting, out of curiosity how the AI training process differs between supervised and unsupervised cases? How does the AI use the amount of error in supervised cases to enhance itself?
Maxime Dolores
@imnikhill10 by backpropagating the error through the network, you might want to follow a simple explanation video on youtube if you're interested in going into the details I'd recommend sentdex Neural Network From Scratch book / series of videos probably the most detailed thing for Deep Neural Networks available For machine learning tasks you could probably find some other good resources online by searching
Carmela Padasas
Got this from SaS Institute: How Artificial Intelligence Works. AI works by combining large amounts of data with fast, iterative processing and intelligent algorithms, allowing the software to learn automatically from patterns or features in the data.