8 Rhim & Park
(“”). They have become familiar to us after AlphaGo. However, its history dates
Neural Network (“ANN”) which emulates the human- neural networks. McCulloch
and Pitts (1943)
proposed the first mathematical model of the ANN. Using the
McCulloch-Pitts model, Frank Rosenblatt (1957)
invented the ANN-Algorithm
(‘Perceptron’). ANN and Perceptron created the concept of Machine Learning which
classified into Rule-based and Learning-based (Machine learning). These two
methodologies have advantages and disadvantages, respectively.
other than ANN, but ANN is
the representative algorithm. ANN created a concept called Machine Learning and
published a monumental paper
with the support of Canada. This paper revealed
that the chronic issues of the ANN could be resolved through pre-training of data
and marked a new milestone in the research on ANN. After this paper, the word,
the Hinton team won with
overwhelming performance at the “
1 W. McCulloch & W. Pitts, A Logical Calculus of the Ideas Immanent in Nervous Activity, 5 BULL. MATH. BIOPHYSICS
2 F. ROSENBLATT, THE PERCEPTRON, A PERCEIVING AND RECOGNIZING AUTOMATON PROJECT PARA 1-29 (Cornell Aeronautical
Laboratory ed., 1957).
3 For example, Regression, Decision Tree, Naive Bayes, Support Vector Machine, etc.
4 G. Hinton, et al., A Fast Learning Algorithm for Deep Belief Nets, 18 NEURAL COMPUTATION 1527-54 (2006).
5 A. Krizhevsky, et al., Imagenet Classification with Deep Convolutional Neural Networks, 25(2) ADVANCES IN NEURAL
INFORMATION PROCESSING SYSTEMS 1097-105 (2012).