Glossary AI Terms

 

Here are some of the most important terms that a student needs to know to understand AI:

Accuracy: The percentage of predictions that a machine learning model makes correctly.

Activation function: A function that defines the output of a node in a neural network given an input or set of inputs. Common activation functions include sigmoid, tanh, ReLU, and softmax

Algorithm: A set of steps a machine can follow to complete a task.

Artificial intelligence (AI): Intelligence demonstrated by machines that can perform tasks that normally require human intelligence, such as reasoning, learning, decision-making, and natural language processing.

Artificial neural network (ANN): See neural network.

Backpropagation: A technique for updating the weights and biases of a neural network based on the error between the predicted output and the actual output. The error is propagated backward through the network, adjusting the parameters in each layer accordingly

Bias: A prejudice in the data that can lead a machine learning model to make inaccurate predictions.

Computer vision: A branch of AI that deals with the analysis and understanding of visual information, such as image processing, face detection, object recognition, scene understanding, and video analysis

Convolutional neural network (CNN): A type of neural network that is especially suited for processing images and other grid-like data. It consists of layers of convolutional filters that extract features from the input data, followed by pooling layers that reduce the dimensionality of the data, and fully connected layers that perform classification or regression tasks

Dataset: A collection of data that is used to train a machine learning model.

Deep learning (DL): A type of ML that uses artificial neural networks to learn from data. A form of machine learning based on neural networks, which are computational models inspired by the structure and function of biological neurons

Ethics: The study of the moral and ethical implications of AI.

Generative adversarial network (GAN): A type of generative model that consists of two neural networks: a generator and a discriminator. The generator tries to produce realistic samples from a latent space, while the discriminator tries to distinguish between real samples from the data distribution and fake samples from the generator. The two networks compete with each other in a minimax game, where the generator tries to fool the discriminator and the discriminator tries to catch the generator. GANs can generate realistic images, videos, text, and audio

Long short-term memory (LSTM): A type of recurrent unit that has a more complex memory state than a simple RNN unit. It has three gates (input, output, and forget) that control how information flows in and out of the memory state. LSTMs can learn long-term dependencies in the data and avoid problems such as vanishing or exploding gradients

Machine learning (ML): A type of AI that allows machines to learn from data without being explicitly programmed. The study of how AI acquires knowledge from training data, using algorithms that can learn from and make predictions on data

Model: A representation of the data that a machine learning model has learned.

Natural language processing (NLP): A branch of AI that deals with the interaction between computers and human languages, such as speech recognition, natural language understanding, natural language generation, machine translation, sentiment analysis, and text summarization

Natural language processing: A field of AI that focuses on developing algorithms that can understand and generate human language.

Neural network: A network of interconnected nodes that process information by passing signals between layers. Each node has an activation function that determines the output of the node given an input or set of inputs

Prediction: The output of a machine learning model when it is given new data.

Recurrent neural network (RNN): A type of neural network that is especially suited for processing sequential data, such as text and speech. It consists of layers of recurrent units that have a memory state that can store information from previous inputs. RNNs can handle variable-length inputs and outputs and can model temporal dependencies in the data

Reinforcement learning: A type of machine learning where the algorithm learns from its own actions and feedback from the environment. The algorithm tries to find an optimal policy that maximizes a reward function over time. Examples of reinforcement learning tasks include game playing and robotics

Robotics: A field of AI focusing on developing robots that can perform tasks autonomously.

Supervised learning: A type of machine learning where the algorithm learns from labeled data, i.e., data that has a known output or target variable. The algorithm tries to find a function that maps the input data to the output data. Examples of supervised learning tasks include classification and regression

Supervised learning: A type of ML in which the model is trained on labeled data, where the input and output are known.

Training: The process of feeding data to a machine learning model so that it can learn.

Unsupervised learning: A type of machine learning where the algorithm learns from unlabeled data, i.e., data that has no known output or target variable. The algorithm tries to find patterns or structures in the data without any guidance. Examples of unsupervised learning tasks include clustering and dimensionality reduction

Unsupervised learning: A type of ML in which the model is trained on unlabeled data, and the model must learn to find patterns in the data on its own.

NFT Vocabulary

NFT marketplace: A digital platform where NFTs can be bought, sold, and traded.

Smart contract: A self-executing contract with the terms of the agreement between buyer and seller directly written into lines of code. Smart contracts enable secure, transparent, and trustless transactions of NFTs.

Airdrop: A free distribution of NFTs to a select group of users.

Floor price: The lowest price at which an NFT is currently being sold on a marketplace.

Gas war: A situation where multiple users are competing to have their transactions processed first, leading to higher gas fees.

Holder: Someone who owns an NFT.

Metadata: Data that describes an NFT, such as its name, description, and image.

PFP: Profile picture, often used to refer to NFT avatars that are used on social media and other online platforms.

Rarity: The scarcity of an NFT, which can affect its value.

Royalty: A percentage of the sale price of an NFT that is paid to the creator or previous owner each time it is sold.

Sniping: Buying an NFT at the last second for a lower price than it is currently listed for.