English Index
Browse every glossary term alphabetically by its English name.
A
- Accountability (Wikipedia)
- 説明責任
Responsibility for decisions, impacts, oversight, and remedies associated with an AI system. - Action (Wikipedia)
- 行動
A choice made by the agent that can affect the environment and future rewards. - Activation function (Wikipedia)
- 活性化関数
A function that transforms a neuron's combined input and introduces nonlinear behavior. - Actor–critic (Wikipedia)
- アクタークリティック
A method combining an actor that updates the policy with a critic that evaluates the actor using value estimates. - Adversarial example (Wikipedia)
- 敵対的サンプル
An input deliberately modified to cause a model to make an incorrect prediction. - Adversarial machine learning (Wikipedia)
- 敵対的機械学習
The study of attacks on machine-learning systems and defenses against them. - Agent (Wikipedia)
- エージェント
The decision-making entity that observes the environment, chooses actions, and receives rewards. - AI agent (Wikipedia)
- AIエージェント
An AI system that observes context, chooses actions, and works toward a goal. - AI ethics (Wikipedia)
- AI倫理
The study and practice of addressing moral issues arising from AI design, deployment, and use. - AI governance (Wikipedia)
- AIガバナンス
The structures, responsibilities, and processes used to direct and oversee AI systems. - AI in finance (Wikipedia)
- 金融分野のAI
The application of AI to financial analysis, decisions, services, and risk management. - AI in healthcare (Wikipedia)
- 医療分野のAI
The application of AI to clinical care, medical operations, research, and public health. - AI in robotics (Wikipedia)
- ロボティクス分野のAI
The use of AI for robot perception, decision-making, learning, and control. - AI regulation and policy (Wikipedia)
- AI規制と政策
Rules and public policies governing the development, deployment, and use of AI. - AI safety and evaluation (Wikipedia)
- AIの安全性と評価
The practice of measuring AI capabilities, limitations, and risks and reducing the likelihood or severity of harmful outcomes. - Algorithm (Wikipedia)
- アルゴリズム
A defined procedure used to solve a problem or perform a computation. - Algorithmic bias (Wikipedia)
- アルゴリズム偏り
Systematic and unfair differences in algorithmic outcomes affecting people or groups. - Algorithmic impact assessment (Wikipedia)
- アルゴリズム影響評価
A structured assessment of an automated system's expected effects, risks, and safeguards. - Anomaly detection (Wikipedia)
- 異常検知
Identifying unusual patterns or outliers in data. - Artificial intelligence (Wikipedia)
- 人工知能
The field concerned with building machines that perform tasks associated with human intelligence. - Artificial neuron (Wikipedia)
- 人工ニューロン
A mathematical unit that combines inputs and applies an activation function to produce an output. - Attention (Wikipedia)
- 注意機構
A mechanism that assigns different weights to input elements when computing a representation. - Automation bias (Wikipedia)
- 自動化バイアス
The tendency to favor suggestions from automated systems even when contrary evidence exists. - Autonomous system (Wikipedia)
- 自律システム
A system that can operate and make decisions with limited direct human control.
B
- Backpropagation (Wikipedia)
- 誤差逆伝播法
An algorithm that computes gradients by propagating output error backward through a neural network. - Bayesian learning (Wikipedia)
- ベイズ学習
Learning that applies Bayes' theorem to update uncertainty about models or parameters. - Benchmark (Wikipedia)
- ベンチマーク
A standardized task, dataset, or measurement used to compare systems. - Binary classification (Wikipedia)
- 二値分類
Classification in which each example is assigned to one of two classes.
C
- Class label (Wikipedia)
- クラスラベル
The category name or identifier predicted in a classification task. - Classification (Wikipedia)
- 分類
The task of assigning an input to one or more discrete categories. - Clinical decision support (Wikipedia)
- 臨床意思決定支援
Software that supplies patient-specific information to assist healthcare decisions. - Clustering (Wikipedia)
- クラスタリング
Grouping similar data points according to their features. - Computer vision (Wikipedia)
- コンピュータビジョン
The field that enables computers to extract information from images and video. - Context window (Wikipedia)
- コンテキストウィンドウ
The amount of input and generated content a model can consider in one interaction. - Continuous value (Wikipedia)
- 連続値
A numerical value that can vary across an interval rather than belonging to a discrete class. - Contrastive learning (Wikipedia)
- 対照学習
Learning representations by bringing related examples closer and separating unrelated examples. - Credit scoring (Wikipedia)
- 信用スコアリング
Estimating the likelihood that a borrower will repay debt. - Cross-validation (Wikipedia)
- 交差検証
A resampling method that evaluates a model across multiple training and validation partitions.
D
- Data (Wikipedia)
- データ
Recorded observations, measurements, symbols, or examples used for analysis and learning. - Data and datasets (Wikipedia)
- データとデータセット
The examples and organized collections used for analysis and machine learning. - Data drift (Wikipedia)
- データドリフト
A change over time in the statistical properties of data received by a deployed system. - Data leakage (Wikipedia)
- データリーク
The unintended use of information during training that would not be available when the model is deployed. - Data poisoning (Wikipedia)
- データポイズニング
An attack that corrupts training data to influence a model's learned behavior. - Data split (Wikipedia)
- データ分割
The process of partitioning a dataset into distinct subsets for model development and evaluation. - Data visualization (Wikipedia)
- データ可視化
Representing data graphically to reveal patterns, trends, and insights. - Dataset (Wikipedia)
- データセット
An organized collection of related data used for analysis or model development. - Deep learning (Wikipedia)
- ディープラーニング
Machine learning based on neural networks with multiple learned representation layers. - Deep reinforcement learning (Wikipedia)
- 深層強化学習
Reinforcement learning that uses deep neural networks to represent policies, value functions, or environment models. - Dimensionality reduction (Wikipedia)
- 次元削減
Reducing the number of features while retaining essential structure or information. - Discount factor (Wikipedia)
- 割引率
A value between zero and one that controls how strongly future rewards affect present decisions. - Downstream task (Wikipedia)
- 下流タスク
A specific application for which a pretrained model is adapted or evaluated.
E
- Embedding (Wikipedia)
- 埋め込み
A learned numerical representation that places items in a vector space. - Environment (Wikipedia)
- 環境
Everything outside the agent that responds to its actions and supplies states and rewards. - Episode (Wikipedia)
- エピソード
One complete sequence of interaction from an initial state until a terminal state. - Evaluation metric (Wikipedia)
- 評価指標
A numerical measure used to assess a model's predictions or behavior. - Explainable AI (Wikipedia)
- 説明可能なAI
Methods and practices intended to make AI behavior understandable to people. - Exploitation (Wikipedia)
- 活用
Choosing actions currently believed to produce the highest return. - Exploration (Wikipedia)
- 探索
Trying actions whose outcomes are uncertain in order to discover potentially better behavior.
F
- Fairness (Wikipedia)
- 公平性
The study and management of unjustified differences in how AI systems affect people or groups. - Feature (Wikipedia)
- 特徴量
A measurable input property supplied to a machine-learning model. - Feature engineering (Wikipedia)
- 特徴量エンジニアリング
The process of creating or transforming model inputs to improve learning and performance. - Feature extraction (Wikipedia)
- 特徴抽出
Transforming raw data into informative derived features. - Feature importance (Wikipedia)
- 特徴量重要度
A measure of how strongly an input feature influences a model's predictions. - Feature selection (Wikipedia)
- 特徴選択
Selecting a useful subset of available input features. - Federated client (Wikipedia)
- 連合クライアント
A participating device or organization that performs local training in federated learning. - Federated learning (Wikipedia)
- 連合学習
Distributed learning in which participants train locally and share model updates rather than raw data. - Few-shot learning (Wikipedia)
- 少数ショット学習
Learning or adapting from only a small number of labeled examples. - Fine-tuning (Wikipedia)
- ファインチューニング
Further training a pretrained model on task-specific data or objectives. - Foundation model (Wikipedia)
- 基盤モデル
A model trained broadly at scale that can be adapted to many downstream tasks. - Fraud detection (Wikipedia)
- 不正検知
Identifying transactions or behavior that may involve deception or unauthorized activity.
G
- Generalization (Wikipedia)
- 一般化
A model’s ability to perform well on data it did not see during training. - Generative model (Wikipedia)
- 生成モデル
A model that learns a data distribution and can produce new samples resembling its training data. - Gradient descent (Wikipedia)
- 勾配降下法
An iterative optimization method that moves parameters in the direction that reduces an objective. - Graph (Wikipedia)
- グラフ
A structure consisting of entities represented as vertices and relationships represented as edges. - Graph neural network (Wikipedia)
- グラフニューラルネットワーク
A neural network designed to learn from nodes, edges, and relationships in graph-structured data. - Grounding (Wikipedia)
- グラウンディング
Connecting a model's output to supplied evidence, data, or observable context.
H
- Hallucination (Wikipedia)
- ハルシネーション
Generated content that is unsupported, false, or inconsistent with the supplied evidence. - 隠れ層
A neural-network layer located between the input and output layers. - Human in the loop (Wikipedia)
- ヒューマン・イン・ザ・ループ
A design in which people provide oversight, decisions, labels, or feedback within an automated process. - Human–AI interaction (Wikipedia)
- 人間とAIの相互作用
The study and design of communication, cooperation, and control between people and AI systems.
I
- Image classification (Wikipedia)
- 画像分類
Assigning a label or category to an image. - Image segmentation (Wikipedia)
- 画像セグメンテーション
The task of assigning image pixels to meaningful regions or classes. - Inference (Wikipedia)
- 推論
Using a trained model to produce an output from new input. - Information retrieval (Wikipedia)
- 情報検索
Finding relevant documents or records in response to an information need. - Input data (Wikipedia)
- 入力データ
The raw information or features supplied to a model. - Intelligent agent (Wikipedia)
- 知的エージェント
A system that perceives an environment and acts to pursue objectives. - Interpretability (Wikipedia)
- 解釈可能性
The degree to which a person can understand how a model produces an output.
L
- Label (Wikipedia)
- ラベル
A known category or target attached to an example in labeled data. - Labeled dataset (Wikipedia)
- ラベル付きデータセット
A collection in which each input is paired with its expected output or target. - Language model (Wikipedia)
- 言語モデル
A model that assigns probabilities to sequences of language elements or predicts language content. - Large language model (Wikipedia)
- 大規模言語モデル
A language model with many parameters trained on large text collections for broad language capabilities. - Latent variable (Wikipedia)
- 潜在変数
An unobserved variable used by a model to explain patterns in observed data. - Localization (Wikipedia)
- 自己位置推定
Estimating the position and orientation of a robot or autonomous system. - Loss function (Wikipedia)
- 損失関数
A formula that measures the difference between predicted and actual outputs.
M
- Machine learning (Wikipedia)
- 機械学習
A field of AI in which systems learn patterns or behavior from data and experience. - Market segmentation (Wikipedia)
- 市場セグメンテーション
Dividing a market into groups with shared characteristics or behavior. - Markov Decision Process (MDP) (Wikipedia)
- マルコフ決定過程
A mathematical framework for sequential decisions defined by states, actions, transition probabilities, rewards, and a discount factor. - Medical imaging (Wikipedia)
- 医用画像
Techniques for creating and interpreting images of the body for clinical purposes. - Memory (Wikipedia)
- メモリ
Stored information that an agent can retain and use across steps or interactions. - Message passing (Wikipedia)
- メッセージパッシング
A graph-learning procedure in which nodes update their representations using information from neighboring nodes. - Meta-learning (Wikipedia)
- メタラーニング
Learning methods that improve a system's ability to learn new tasks from limited experience. - Modality (Wikipedia)
- モダリティ
A particular form or channel of information, such as text, vision, or audio. - Model (Wikipedia)
- モデル
A mathematical representation that processes input data to produce predictions, classifications, or decisions. - Model aggregation (Wikipedia)
- モデル集約
The process of combining model updates received from multiple federated clients. - Model deployment (Wikipedia)
- モデル展開
Making a trained model available in a production system or application. - Model evaluation (Wikipedia)
- モデル評価
The systematic measurement of how well a model performs for its intended purpose. - Model monitoring (Wikipedia)
- モデル監視
Tracking a deployed model's behavior, performance, inputs, and operational health. - Model-based reinforcement learning (Wikipedia)
- モデルベース強化学習
Reinforcement learning that uses a learned or supplied model of environment dynamics to plan or improve behavior. - Model-free reinforcement learning (Wikipedia)
- モデルフリー強化学習
Reinforcement learning that learns values or policies directly from experience without learning a model for planning. - Motion planning (Wikipedia)
- 動作計画
Computing a feasible sequence of movements from a starting configuration to a goal. - Multiclass classification (Wikipedia)
- 多クラス分類
Classification in which an example is selected from more than two possible classes. - Multimodal AI (Wikipedia)
- マルチモーダルAI
AI that processes or combines more than one kind of data, such as text, images, audio, or video. - Multimodal embedding (Wikipedia)
- マルチモーダル埋め込み
A representation that places information from different modalities in a shared or aligned vector space.
N
- Named-entity recognition (Wikipedia)
- 固有表現抽出
The task of locating and categorizing names such as people, organizations, and places in text. - Natural language processing (Wikipedia)
- 自然言語処理
The field concerned with computational processing and generation of human language. - Neural network (Wikipedia)
- ニューラルネットワーク
A model composed of connected processing units arranged in layers.
O
- Object detection (Wikipedia)
- 物体検出
The task of locating and classifying objects within an image or video. - Objective function (Wikipedia)
- 目的関数
A numerical function that an optimization process seeks to minimize or maximize. - Optimization algorithm (Wikipedia)
- 最適化アルゴリズム
A procedure for finding parameter values that improve or minimize an objective. - Output label (Wikipedia)
- 出力ラベル
The expected category or value associated with an input.
P
- Planning (Wikipedia)
- 計画
Selecting and ordering actions intended to move from a current state toward a goal. - Policy (Wikipedia)
- 方策
A rule or probability distribution that determines which action an agent selects in each state. - Policy gradient (Wikipedia)
- 方策勾配
A family of policy-based methods that improves a policy by following the gradient of expected return. - Policy-based method (Wikipedia)
- 方策ベース手法
A method that optimizes a policy directly instead of deriving it only from learned values. - Posterior probability (Wikipedia)
- 事後確率
A probability updated after combining prior belief with observed evidence. - Prediction (Wikipedia)
- 予測
An output produced by a trained model. - Prediction error (Wikipedia)
- 予測誤差
The difference between a predicted value and its target value. - Pretext task (Wikipedia)
- 代理タスク
A task constructed from unlabeled data to provide a learning signal for representation learning. - Pretrained model (Wikipedia)
- 事前学習済みモデル
A model whose parameters were learned before adaptation to the current task. - Pretraining (Wikipedia)
- 事前学習
Initial training on broad data before a model is adapted for a more specific purpose. - Prior probability (Wikipedia)
- 事前確率
A probability expressing belief before incorporating the current evidence. - Probabilistic model (Wikipedia)
- 確率モデル
A model that represents variables and outcomes using probability distributions. - Probability distribution (Wikipedia)
- 確率分布
A mathematical description of the probabilities assigned to possible outcomes.
Q
- Q-learning (Wikipedia)
- Q学習
A model-free, value-based algorithm that learns the expected return for taking each action in each state.
R
- Random variable (Wikipedia)
- 確率変数
A variable whose possible numerical values are outcomes of a random process. - Red teaming (Wikipedia)
- レッドチーミング
Adversarial testing that searches deliberately for failures, vulnerabilities, or harmful behavior. - Regression (Wikipedia)
- 回帰
A supervised-learning task that estimates a continuous numerical target. - Reinforcement learning (Wikipedia)
- 強化学習
A machine-learning approach in which an agent learns behavior through interaction and feedback from rewards. - Representation learning (Wikipedia)
- 表現学習
Learning useful features or representations directly from data. - Retrieval-augmented generation (Wikipedia)
- 検索拡張生成
A generation method that retrieves external information and supplies it to a generative model as context. - Return (Wikipedia)
- 収益
The total reward accumulated from a time step onward, usually with future rewards discounted. - Reward (Wikipedia)
- 報酬
A numerical feedback signal that indicates the immediate desirability of an outcome. - Robot perception (Wikipedia)
- ロボット知覚
The process by which a robot interprets sensor data to understand itself and its surroundings. - Robustness (Wikipedia)
- 頑健性
The ability of a system to maintain acceptable behavior under variation, noise, or attack.
S
- Sampling (Wikipedia)
- サンプリング
The process of drawing examples from a probability distribution or generative model. - Self-attention (Wikipedia)
- 自己注意機構
Attention in which elements of one sequence attend to other elements in that same sequence. - Self-supervised learning (Wikipedia)
- 自己教師あり学習
Learning in which supervisory signals are derived from the structure of the input data itself. - Semantic similarity (Wikipedia)
- 意味的類似度
A measure of how closely two items match in meaning rather than surface form. - Sensor fusion (Wikipedia)
- センサフュージョン
Combining measurements from multiple sensors to obtain a more reliable estimate. - Spam email detection (Wikipedia)
- スパムメール検出
Classifying email as spam or not spam. - State (Wikipedia)
- 状態
A representation of the information needed to describe the environment at a particular time. - State transition (Wikipedia)
- 状態遷移
The change from one state to another after the agent takes an action. - Supervised learning (Wikipedia)
- 教師あり学習
A machine-learning approach in which a model is trained on labeled examples. - Support set (Wikipedia)
- サポートセット
The small labeled example set supplied for adaptation in a few-shot learning task.
T
- Target task (Wikipedia)
- 対象タスク
The task to which transferred knowledge or a pretrained model is being applied. - Target value (Wikipedia)
- 目標値
The expected output used in supervised learning and absent from unsupervised training data. - Target variable (Wikipedia)
- 目的変数
The outcome that a supervised model is trained to predict. - Terminal state (Wikipedia)
- 終端状態
A state that ends an episode and after which no further action is taken in that episode. - Test data (Wikipedia)
- テストデータ
Held-out data used to estimate a finished model's performance on unseen examples. - Time step (Wikipedia)
- 時間ステップ
One discrete point in an interaction sequence at which the agent observes, acts, and receives feedback. - Token (Wikipedia)
- トークン
A unit of text or other input that a model processes as one element of a sequence. - Tokenization (Wikipedia)
- トークン化
The process of dividing input into units that a language-processing system can handle. - Tool use (Wikipedia)
- ツール使用
The ability of an AI system to invoke external functions, software, or services to complete work. - Training (Wikipedia)
- 訓練
The process of adjusting a model’s parameters using data. - Training data (Wikipedia)
- 訓練データ
Examples used to fit a model's parameters or learned behavior. - Training, validation, and test data (Wikipedia)
- 訓練・検証・テストデータ
Separate data subsets used to fit a model, tune choices, and estimate final performance. - Transfer learning (Wikipedia)
- 転移学習
Reusing knowledge learned for one task or domain to improve learning on another. - Transformer (Wikipedia)
- トランスフォーマー
A neural-network architecture that processes relationships between tokens primarily through attention mechanisms.
U
- Unseen data (Wikipedia)
- 未見データ
Data that a model did not encounter during training. - Unsupervised learning (Wikipedia)
- 教師なし学習
A machine-learning approach in which a model learns patterns from unlabeled data.
V
- Validation data (Wikipedia)
- 検証データ
Data used during development to compare settings or select models without fitting their parameters directly. - Value function (Wikipedia)
- 価値関数
A function that estimates the expected return from a state, or from taking an action in a state, under a policy. - Value-based method (Wikipedia)
- 価値ベース手法
A method that learns value estimates and derives action choices from them. - Vector (Wikipedia)
- ベクトル
An ordered collection of numbers used to represent position, direction, or learned features. - Vector database (Wikipedia)
- ベクトルデータベース
A database designed to store embeddings and retrieve nearby vectors efficiently. - Vision-language model (Wikipedia)
- 視覚言語モデル
A model trained to represent and reason across visual and linguistic information.