Glossary
AI_related acronyms and terms and their meaning.
AGI: Artificial General Intelligence - A hypothetical machine that has the ability to understand or learn any intellectual task that a human being can.
AI: Artificial Intelligence - The simulation of human intelligence in machines that are programmed to think like humans and mimic their actions.
ANN: Artificial Neural Network - A computational model based on the structure and functions of biological neural networks.
API: Application Programming Interface - A set of protocols for integrating software applications.
AUC: Area Under the Curve - A metric used to evaluate the performance of a binary classification model.
AR: Augmented Reality - An interactive experience of a real-world environment where the objects that reside in the real world are enhanced by computer-generated perceptual information.
BERT: Bidirectional Encoder Representations from Transformers - A pre-trained transformer-based language model that can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of NLP tasks.
CART: Classification and Regression Trees - A decision tree algorithm used for classification and regression analysis.
CNN: Convolutional Neural Network - A type of neural network that is often used in image recognition and processing.
CRF: Conditional Random Field - A probabilistic graphical model used for structured prediction tasks in natural language processing and computer vision.
DBN: Deep Belief Network - A generative graphical model composed of multiple layers of latent variables with connections between them.
DL: Deep Learning - A subset of machine learning that involves training large and deep neural networks to learn from large amounts of data.
DQN: Deep Q-Network - A type of reinforcement learning algorithm used in AI systems to learn how to take actions in an environment in order to maximize a cumulative reward signal.
ELMo: Embeddings from Language Models - A deep contextualized word representation model that uses a bidirectional LSTM architecture trained on a large corpus of text data.
GAN: Generative Adversarial Network - A type of neural network architecture used in unsupervised machine learning that involves two neural networks competing against each other to generate new data.
GPU: Graphics Processing Unit - A specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display device.
HMM: Hidden Markov Model - A statistical Markov model in which the system being modeled is assumed to be a Markov process with unobserved (hidden) states.
LDA: Latent Dirichlet Allocation - A generative statistical model used for topic modeling in natural language processing and machine learning.
LLM: Large Language Model - A Large AI model that uses deep learning techniques on large-scale text input data to generate human-like text at scale based on a given prompt or context.
LSTM: Long Short-Term Memory - A type of RNN (recurrent neural network) architecture used in deep learning for processing sequential data such as natural language or time series data.
MCMC: Markov Chain Monte Carlo - A class of algorithms for sampling from probability distributions based on constructing a Markov chain that has the desired distribution as its equilibrium distribution.
ML: Machine Learning - The study of computer algorithms that improve automatically through experience and by the use of data.
NLP: Natural Language Processing - The ability of computers to understand, interpret, and generate human language.
PR AUC: Precision Recall Area Under the Curve - An evaluation metric used for binary classification problems that considers both precision and recall across different probability thresholds.
RBM: Restricted Boltzmann Machine - A generative stochastic neural network that can learn a probability distribution over its set of inputs.
RL: Reinforcement Learning - A type of machine learning where an agent learns to behave in an environment by performing certain actions and receiving rewards or punishments based on those actions.
RLHF: Reinforcement Learning with Human Feedback - A type of reinforcement learning where humans score interaction results to provide feedback for the RL agent to learn.
RNN: Recurrent Neural Network - A type of neural network designed for processing sequential data such as natural language or time series data.
SVM: Support Vector Machine - A type of supervised learning algorithm used for classification and regression analysis.
VAE: Variational Autoencoder.
Some reinforcement learning-specific acronyms and terms, which may be relevant and helpful.
PG: Policy Gradient - A class of reinforcement learning algorithms that optimize a policy by directly maximizing the expected reward.
AC: Actor-Critic - A class of reinforcement learning algorithms that combine value-based and policy-based methods by using two separate models for estimating value functions and policies respectively.
PPO: Proximal Policy Optimization - A type of policy gradient algorithm that uses a clipped surrogate objective function to prevent large policy updates.
A2C: Advantage Actor-Critic - A type of actor-critic algorithm that uses an advantage function to estimate the advantage of taking a particular action in a particular state compared to the average expected reward for that state.
TRPO: Trust Region Policy Optimization - A type of policy gradient algorithm that constrains policy updates to be within a trust region around the current policy.
TD: Temporal Difference - A class of model-free reinforcement learning methods that learn by bootstrapping from the current estimate of the value function.
MC: Monte Carlo - A class of model-free reinforcement learning methods that learn by averaging over complete episodes of experience.
DDPG: Deep Deterministic Policy Gradient - A type of actor-critic algorithm that uses a deterministic policy instead of a stochastic policy.
HER: Hindsight Experience Replay - A technique used in reinforcement learning that allows an agent to learn from failed experiences by re-framing them as successful experiences with different goals.