Generative AI
Variational autoencoders (VAEs): VAEs are generative models that encode the input data into a latent space, and then decode it back into the output data. The latent space is a lower-dimensional representation of the data that captures its essential features. VAEs can generate new data by sampling from the latent space.
A VAE system works by detecting deviations (anomalies) and either removing these from the generation pipeline or highlighting these deviations to the end user for example if data was being used to detect fraudulent entries.
Diffusion models: Diffusion models are generative models that gradually add noise to the input data until it becomes random, and then reverse the process to reconstruct the output data. Diffusion models can generate new data by starting from random noise and applying the reverse process.
Generative adversarial networks (GANs): GANs are generative models that consist of two competing neural networks: a generator and a discriminator. The generator tries to create fake data that looks like real data, while the discriminator tries to distinguish between the real and fake data. The generator learns from the feedback of the discriminator and improves its output over time.
These opposing systems refine data outputs and overtime can improve accuracy by questioning the results produced.