Types

SVM (Support Vector Machine)

• Used for regression and classification

Transformers

Neural Networks

• Neural networks may have billions of nodes
• As more input data is added, the connections and weights between the nodes are adjusted

Convolucional Neural Networks (CNN)

RNN (Recurrent Neural Network)

• Meant for sequential data such as time-series
• Useful in
• Speech recognition
• Time-series predictions

Multilayer Perceptrons

ResNet (Residual Network)

• Specialized deep CNNs
• Useful for
• Image recognition
• Object detection
• Facial recognition

WaveNet

• Model to generate raw audio waveform
• Used in speech synthesis

GAN (Generative Adversarial Network)

• Models used to generate synthetic data such as images, videos or sounds
• The synthetic data resembles the training data
• Helpful for data augmentation

XGBoost (Extreme Gradient Boosting)

• An implementation of gradient booting
• Used for regressions