An activation function in an artificial neural network (ANN) is a mathematical function that processes a neuron's input and determines its output. It helps decide whether a neuron should be activated, allowing the network to capture complex relationships in data. When non-linear activation functions are used, they enable the network to model non-linear patterns and interactions.
AdaBoost (Adaptive Boosting) is an ensemble learning algorithm that combines multiple weak learners (usually decision stumps) to create a strong classifier.
Adam (Adaptive Moment Estimation) is a popular optimization algorithm in deep learning that combines momentum and adaptive learning rates.
A/B testing is a statistical method used to compare two versions of a model, system, or experiment to determine which one performs better.
The presence of systematic errors in machine learning models due to biased data or assumptions.
Seen in biased hiring algorithms.
AlphaGo is a deep reinforcement learning-based AI developed by DeepMind that defeated human champions in Go.
Used in reinforcement learning research.
Anomaly detection is the process of identifying data points that deviate significantly from the norm.
Used in fraud detection and cybersecurity.
AI is the broad field of creating machines that can mimic human intelligence, including reasoning, learning, and decision-making.
AI-powered assistants like Siri and Alexa.
ANN is a computational model inspired by the structure of the human brain, consisting of layers of interconnected neurons that process data and learn patterns.
Used in image recognition and self-driving cars.
A technique that allows neural networks to focus on important parts of input data, widely used in NLP.
Used in machine translation models like Transformer.
An Autoencoder is a type of neural network used for unsupervised learning that compresses and reconstructs data.
Used in anomaly detection and data compression.
Automated Machine Learning (AutoML) refers to the process of automating ML model selection and hyperparameter tuning.
Used in cloud-based ML platforms.
Data augmentation involves artificially increasing training data by applying transformations like rotation, flipping, and scaling.
Used in image classification to improve generalization.
A statistical model where future values depend on past values.
Used in time series forecasting.
Average pooling is a down-sampling technique in convolutional neural networks (CNNs) that reduces dimensionality by averaging values in a local region.
Used in image processing to retain feature information.
A technique used to find the nearest data points in high-dimensional space efficiently.
Used in recommendation systems.
An optimization technique that adjusts the learning rate dynamically during training.
Used in optimizers like Adam and RMSprop.
A type of memory system in artificial neural networks that recalls stored patterns when given partial input.
Used in pattern recognition and AI assistants.
A statistical method where future values of a variable are predicted based on past values.
Used in time series forecasting.
A technique in machine learning where sampling is adjusted dynamically based on model performance or uncertainty.
Used in active learning models.
The process of selecting the most relevant features in a dataset to improve model performance.
Used in dimensionality reduction and preprocessing.
A technique where small, intentional modifications are made to input data to trick a machine learning model.
Used to test the robustness of deep learning models.
A computational method used to find near-optimal solutions when exact solutions are computationally infeasible.
Used in NP-hard problems like clustering.
A technique used to segment an image by computing different threshold values based on local regions.
Used in image processing and OCR.
A technique used in deep learning to compute gradients efficiently for optimization.
Used in backpropagation during model training.
A structured workflow for automating the entire machine learning process, from data preprocessing to model deployment.
Used in cloud-based ML automation.
A machine learning algorithm that combines decision trees using boosting techniques to improve accuracy.
Used in ensemble learning and Kaggle competitions.
A Bayesian inference method used when likelihood functions are computationally expensive to evaluate.
Used in probabilistic modeling.
A machine learning approach where the model actively selects the most informative data points for labeling.
Used in semi-supervised learning.
A variant of gradient descent that adapts learning rates individually for each parameter.
Used in optimization algorithms like Adagrad.
A neural network framework designed to classify changing data over time without forgetting previous knowledge.
Used in clustering and speech recognition.
A concept in AI ethics that ensures models make unbiased decisions across different demographic groups.
Used in fairness-aware AI systems.
A process where complex data structures and behaviors are represented in simpler terms for better understanding.
Used in knowledge representation and symbolic AI.
A clustering algorithm that identifies exemplars in a dataset and groups data points based on similarity.
Used in image segmentation and recommendation systems.
A computational modeling approach where autonomous agents interact based on predefined rules to simulate complex systems.
Used in social simulations and economic forecasting.
A statistical measure used to evaluate the quality of a model, balancing goodness of fit and complexity.
Used in model selection for regression and classification problems.
The challenge of ensuring that AI systems act in accordance with human values and ethical guidelines.
Used in AI safety and governance research.
An optimization technique in decision tree algorithms that reduces the number of nodes evaluated in minimax search.
Used in AI game playing, such as chess engines.
A technology that overlays digital information onto the real world, often powered by AI for object recognition.
Used in AR apps like Pokémon GO and virtual try-ons.
The process of using AI to automatically extract and generate the most relevant features from raw data.
Used in AutoML frameworks for improved model accuracy.
ML is a subset of AI that enables machines to learn patterns from data and make predictions or decisions without explicit programming.
Spam detection in emails using classification models.
DL is a subset of ML that uses artificial neural networks to process complex data and perform high-level computations.
Image recognition in self-driving cars.
Gen AI refers to AI models that generate new content, including text, images, and code, using trained knowledge bases.
AI models like ChatGPT and Stable Diffusion that generate text and images.