Practical perspective of artificial neural networks as a function approximator, not an almighty black box

Tae-Hyung Kim · 2014

In the realm of Computational Intelligence/Machine Learning (CI/ML), artificial neural networks (ANNs) are believed to have potential to mimic human brain. As a researcher, maintaining this dream is important. However, one should realize an artificial neuron is a model, not an actual biological neuron. This paper indicates the logical leap why ANNs are not like human brain.

Read the paper · More papers on PaperTik