Page 10 - IJB-6-1
P. 10
Machine learning in bioprinting
Figure 2. A typical machine learning process.
learning has been used in aircraft control, robot Table 1. Main specific techniques in each machine
motion control, traffic light control, web system learning method
configuration, and games (e.g., AlphaGo). Machine learning Main specific techniques
In addition to these three learning paradigms, methods
some new machine learning methods have Supervised learning Decision trees, logistic regression,
also been developed, such as semi-supervised decision forests, support vector
learning. More details can be found in a review machines, kernel machines,
paper on machine learning [36] . There are many Bayesian classifiers
algorithms available in each machine learning Unsupervised learning k-means, generative adversarial
method; the main specific techniques in each networks, expectation-maximization
algorithm, Hebbian Learning, self-
method are listed in Table 1. Some of these organizing map, adaptive resonance
specific techniques have been applied in 3D theory
printing. In the next section, machine learning Reinforcement learning Monte Carlo, Q-learning, Soft Actor-
used in 3D printing processes will be briefly Critic, proximal policy optimization,
Trust Region Policy Optimization,
reviewed and the corresponding inspirations for Deep Q-Network, deep deterministic
3D bioprinting will be proposed. policy gradient
6 International Journal of Bioprinting (2020)–Volume 6, Issue 1
