按方法、领域与证据探索科学。

Homomorphic Encryption

Homomorphic Encryption (HE) is a cryptographic framework that allows arbitrary computations to be performed directly on encrypted data without requiring decryption. First realized as a fully general construction by Craig Gentry in 2009 using ideal lattices, it enables a server to process sensitive data and return an en

computer vision

Hough Transform

The Hough Transform is a technique for detecting lines, circles, and other geometric shapes in digital images. Originally patented by Paul Hough in 1962 and popularized in computer vision by Duda and Hart in 1972, the Hough Transform converts edge points in image space to curves in a parameter space (accumulator space)

linguistics

HPSG

Head-Driven Phrase Structure Grammar (HPSG) is a constraint-based grammatical framework developed by Carl Pollard and Ivan Sag in 1987. HPSG represents linguistic information (phonological, syntactic, semantic) in typed feature structures and derives well-formed expressions through constraints on these structures. Unli

biomechanics

Hydrogel Rheology

Hydrogel rheology characterizes the mechanical viscoelastic properties of hydrogels used in tissue engineering, drug delivery, and biomedical devices. By measuring storage modulus (elastic component), loss modulus (viscous component), and their frequency dependence, practitioners assess gel stiffness, degradation, and

geoscience

Hydrogeological Survey

Hydrogeological survey is the systematic characterization of groundwater systems, including aquifer geometry, water quality, flow paths, and recharge-discharge dynamics. Rooted in Darcy's law (1856) and quantified by Theis (1935), this method is essential for water resource management, contaminant remediation, and haza

oceanography

Hydrothermal Plume Mapping

Hydrothermal plume mapping is an integrated method for detecting, characterizing, and tracking buoyant plumes of hot, mineral-rich water discharged from submarine hydrothermal vents on the seafloor. Developed by Ed Baker and colleagues in the 1980s, hydrothermal plume mapping combines temperature, conductivity, optical

meteorology

HYSPLIT

HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory Model) is a widely used atmospheric transport and dispersion model developed by NOAA's Air Resources Laboratory. It computes air parcel trajectories and pollutant dispersion using Lagrangian tracking to simulate how contaminants and particles move through

genetics

IBD Mapping

Identity-by-descent (IBD) mapping is a genetic mapping technique that identifies disease loci in consanguineous families or isolated populations by detecting homozygous chromosomal segments shared among affected individuals. Developed by Lander and Botstein in 1987, this method exploits the fact that rare disease allel

deep learning

Image Classification

Image classification is the task of assigning a single semantic label to an entire image from a fixed set of categories. Modern approaches rely on deep convolutional neural networks (CNNs) or Vision Transformers (ViTs) trained end-to-end on large labeled datasets such as ImageNet, achieving superhuman accuracy on many

computer vision

Image Morphology Operations

Morphological image processing, introduced by Jean Serra in 1982, is a technique based on set theory that reshapes and analyzes image regions using geometric structuring elements. Core operations include erosion and dilation, which can be combined into more complex operations like opening and closing, enabling noise re

acoustics

Impedance Tube

An impedance tube (or Kundt tube) is a laboratory apparatus for measuring the acoustic absorption coefficient and surface impedance of materials. Originally developed by August Kundt in 1866, the technique has been standardized by ASTM and ISO for characterizing noise-control and acoustic-treatment materials. The imped

text mining

Implicit Sentiment Analysis

Implicit sentiment analysis detects indirect, context-dependent sentiment in text where no explicit opinion word is present — such as irony, metaphor, or understated criticism. Unlike standard sentiment analysis, which relies on surface-level polarity signals, this method interprets meaning from surrounding context, pr

simulation

Importance Sampling

Importance sampling is a Monte Carlo variance-reduction technique that shifts the sampling distribution toward the region of interest — typically a rare or extreme event — so that informative samples are drawn far more often than under the original distribution. Developed at the RAND Corporation by Herman Kahn and Theo

deep learning

Inception Network

The Inception Network, introduced by Szegedy et al. at Google in 2015 and submitted to CVPR under the name GoogLeNet, is a 22-layer deep convolutional neural network designed for large-scale image recognition. Its defining contribution is the Inception module, which applies convolutions of multiple kernel sizes in para

civil engineering

Incremental Dynamic Analysis

Incremental dynamic analysis (IDA) is a method that runs time-history analyses on a structure with a single ground motion record, progressively increasing the intensity until the structure reaches a specified performance level or collapses. Introduced by Vamvatsikos and Cornell in 2002, this approach efficiently genera

machine learning

Independent Component Analysis

Independent Component Analysis (ICA) is a computational method for separating a multivariate signal into additive, statistically independent subcomponents. Formalized by Pierre Comon in 1994, ICA became the foundational framework for blind source separation and is widely applied in neuroimaging (fMRI, EEG), speech proc

applied physics

Independent Vector Analysis

Independent Vector Analysis (IVA) is a multivariate extension of Independent Component Analysis that jointly separates multiple datasets while maintaining dependencies within each dataset. Developed by Lee, Lewicki, and Sejnowski in the 2000s, IVA is used for blind source separation in multi-channel audio, brain imagin

model evaluation

Inertia (Within-Cluster Sum of Squares)

Inertia, also called Within-Cluster Sum of Squares (WCSS), is a measure of cluster cohesion that quantifies how tightly points are grouped around their cluster centroids. Lower values indicate more compact, cohesive clusters. Inertia is the primary objective function for k-means clustering and has been a fundamental me

text mining

Information Extraction

Information extraction (IE) is a natural-language-processing task that converts unstructured text into structured information — such as events, relations, and attributes — so that facts buried in free-form documents become machine-readable records. The task was consolidated in early surveys by Cowie and Lehnert (1996)

deep learning

Informer

Informer is a Transformer-based model introduced by Zhou et al. in 2021 for long-sequence time-series forecasting, using a ProbSparse self-attention mechanism that lowers the computational complexity of the standard Transformer to O(L log L). It is built for problems that demand predictions across thousands of future s

aerospace

INS Error Model

The INS Error Model is a mathematical framework that characterizes how errors in inertial sensor measurements propagate through a navigation system's estimates of position, velocity, and attitude. Developed during the 1960s and refined through decades of navigation research, the error model enables design of optimal es

geophysics

InSAR

Interferometric Synthetic Aperture Radar (InSAR) is a radar remote sensing technique that measures millimeter-scale ground surface deformation by analyzing the phase difference between radar images acquired from slightly different orbital positions. Pioneered by Gabriel, Goldstein, and Zebker in 1989, InSAR has become

deep learning

Instance Segmentation

Instance segmentation is a computer vision task that simultaneously detects every distinct object in an image and produces a precise pixel-level mask for each individual object instance. Unlike semantic segmentation, which labels every pixel with a class, instance segmentation distinguishes between separate objects of

music information retrieval

Instrument Recognition

Instrument recognition is the task of automatically identifying which musical instruments are present in an audio recording. Formalized by Eronen et al. (2005), it addresses timbre—the tonal quality distinguishing one instrument from another. Instrument recognition is essential for music analysis, transcription, automa

biomechanics

Integrate-and-Fire Model

The integrate-and-fire (IF) model is a simplified neuronal model that captures spike generation by integrating synaptic inputs until membrane potential reaches a threshold, at which point a spike is emitted. First proposed by Louis Lapicque in 1907 and refined with leak (leaky integrate-and-fire, LIF), it remains a sta

text mining

Intent Detection

Intent detection is a natural-language-understanding task that classifies the purpose behind a user utterance — such as making a reservation, asking for information, or filing a complaint — into one of a set of predefined intent classes. It is a core NLU component of conversational interfaces and customer-service autom

linguistics

Internal Reconstruction

Internal Reconstruction is a historical linguistic method that reconstructs earlier stages of a single language by identifying internal inconsistencies, morphological irregularities, and distributional patterns within the language itself. Unlike the Comparative Method, which relies on comparing related languages, Inter

cryptography

Intrusion Detection System

An Intrusion Detection System (IDS) is a security tool that monitors network traffic and system activity to identify unauthorized access attempts, malware infections, and policy violations. Introduced by Dorothy Denning in 1987, IDS employs two main detection paradigms: signature-based (matching known attack patterns)

biomechanics

Inverse Dynamics

Inverse dynamics is a biomechanical analysis technique that estimates the forces and moments acting on joints during movement by working backward from observed motion and ground reaction forces. Introduced by David Winter in the early 1990s, it is fundamental to understanding how muscles and joints generate and control

manufacturing

Inverse Kinematics

Inverse kinematics is the computational problem of determining the joint angles required to position and orient the end-effector (tool) of an articulated mechanism at a desired pose (position and orientation). In contrast to forward kinematics, which computes end-effector position from joint angles, inverse kinematics

materials science

Ising Model Monte Carlo

Ising Model Monte Carlo simulation is a computational method for studying phase transitions and magnetic ordering in materials by stochastically sampling configurations of binary spins on a lattice. Originating from Ernst Ising's 1925 theoretical model and combined with Metropolis algorithm in 1953, Ising Monte Carlo e

machine learning

Isolation Forest

Isolation Forest is an unsupervised machine-learning method for anomaly and outlier detection, introduced by Liu, Ting and Zhou in 2008, that isolates anomalies through random partitioning of the data. It works without any labelled anomaly data and scales to high-dimensional datasets.

machine learning

Isomap

Isomap (Isometric Feature Mapping) is a manifold learning algorithm introduced by Tenenbaum, de Silva, and Langford in 2000 that discovers the intrinsic low-dimensional geometry of high-dimensional data by preserving geodesic — rather than straight-line Euclidean — distances between all pairs of points. It was one of t

control theory

Iterative Learning Control

Iterative Learning Control (ILC) is a control method for systems that perform the same task repeatedly (trajectory tracking over a fixed time interval). The key idea is to use error information from previous trials to update the input for the next trial, progressively improving tracking accuracy. Pioneered by Arimoto e

deep learning

iTransformer

iTransformer is a deep-learning architecture for multivariate time-series forecasting introduced by Liu et al. at ICLR 2024. Its defining idea is to invert the conventional Transformer tokenisation strategy: instead of treating each time step as a token, iTransformer treats each variate (sensor channel or feature serie

model evaluation

Jaccard Index

The Jaccard index measures the similarity between predicted and true label sets by computing the ratio of intersection to union. It is widely used in multi-label classification and set-based similarity tasks where partial overlap is important.

biomechanics

Joint Reaction Force

Joint reaction force (JRF) estimation calculates the contact forces transmitted across joints during movement using inverse dynamics combined with anatomical modeling. First validated in vivo by Bergmann and colleagues using instrumented hip implants, JRF estimation is essential for understanding joint degeneration, de

privacy

k-Anonymity

k-Anonymity is a formal privacy model introduced by Latanya Sweeney in 2002 to protect individuals when personal data is released for research or public use. It requires that every record in a published dataset be indistinguishable from at least k−1 other records with respect to a designated set of quasi-identifying at

network analysis

k-Core Decomposition

k-Core Decomposition is a graph-theoretic method that partitions the vertices of a network into a nested sequence of subgraphs called k-cores. A k-core is the maximal subgraph in which every vertex has at least k neighbors within that subgraph. Introduced by Stephen B. Seidman in 1983, the method assigns each vertex a

machine learning

K-means

K-means is a classic unsupervised partitional clustering algorithm that divides a dataset into K non-overlapping groups by iteratively assigning each observation to its nearest centroid and updating centroids as the mean of their assigned points. It is one of the most widely used exploratory tools in machine learning a

machine learning

K-Means Clustering

K-Means Clustering is a centroid-based partitional clustering algorithm, traced to J. MacQueen in 1967, that splits data into k clusters by assigning each observation to its nearest cluster centre. It is widely used for marketing segmentation, customer grouping, and exploratory analysis.

machine learning

K-Nearest Neighbors

K-Nearest Neighbors (KNN), formalized by Cover and Hart in 1967, is a non-parametric, instance-based method that classifies or predicts a new observation by looking at the k closest examples in the training data. For classification it takes a majority vote among those neighbors; for regression it averages their values.

signal processing

Kalman Filter for Signal Tracking

The Kalman filter is a recursive algorithm that optimally estimates the state of a linear dynamic system from noisy measurements, minimizing mean-square error. Introduced by Rudolf Kalman in 1960, it revolutionized control theory, navigation, and signal processing by enabling real-time optimal estimation for time-varyi

machine learning

Kernel PCA

Kernel Principal Component Analysis (Kernel PCA) is a nonlinear dimensionality-reduction method introduced by Bernhard Schölkopf, Alexander Smola, and Klaus-Robert Müller in 1997–1998. It extends classical linear PCA to curved, non-linear data manifolds by implicitly mapping input data into a high-dimensional feature s

text mining

Keyword Extraction

Keyword extraction is a natural-language-processing task that automatically identifies the words or phrases that best represent the content of a document. It turns a body of free text into a compact, ranked list of key terms, drawing on statistical, graph-based methods such as TextRank (Mihalcea & Tarau, 2004), or embe

astronomy

Kinematic Distance

Kinematic distance is a method for estimating distances to objects in the Milky Way using their observed radial velocities and the known rotation curve of the Galaxy. Developed in the 1950s by Bert Westerhout and others, this technique enables distance determination to distant molecular clouds and masers without trigon

quantum computing

KKR Method

The Korringa-Kohn-Rostoker (KKR) method is a powerful multiple-scattering approach for calculating electronic band structures and properties of periodic and disordered solids. Developed in the late 1940s, KKR treats electrons as scattering from atomic potentials in a muffin-tin geometry, enabling efficient calculations

deep learning

Knowledge Distillation

Knowledge Distillation is a model-compression technique, introduced by Geoffrey Hinton and colleagues in 2015, that trains a small student model using the soft-label outputs of a large teacher model. Distilled models such as DistilBERT and TinyBERT reach roughly 97% of the larger model's performance while running far f

network analysis

Knowledge Graph Analysis

Knowledge Graph Analysis is a framework for representing, storing, and reasoning over structured factual knowledge as a directed graph of entities and typed relations. Entities (nodes) and relationships (edges) are expressed as subject–predicate–object triples, enabling rich querying, inference, and integration of hete

text mining

Knowledge Graph Construction

Knowledge graph construction is a text-mining pipeline that turns unstructured text into a structured graph of entities and the relations between them. Drawing on the synthesis of Hogan et al. (2021) and the relational-machine-learning review of Nickel et al. (2016), it represents knowledge as nodes (entities such as p

network analysis

Knowledge Graph Embeddings

Knowledge Graph Embeddings (KGE) are a family of methods that represent entities and relations in a knowledge graph as dense, low-dimensional vectors in a continuous space. The foundational model, TransE, was introduced by Bordes, Usunier, García-Durán, Weston, and Yakhnenko in 2013. TransE treats each relation as a tr

meteorology

Kohler Theory

Köhler theory is a foundational framework in cloud microphysics that predicts the equilibrium supersaturation required for an aerosol particle of given size and composition to grow into a cloud droplet. Published in 1936 by Hilding Köhler, it combines the Kelvin effect (vapor pressure enhancement over curved surfaces)

deep learning

Kolmogorov-Arnold Networks

Kolmogorov-Arnold Networks (KAN) is a neural network architecture introduced by Liu et al. in 2024 that replaces linear transformations with learned univariate functions on edges. Inspired by the Kolmogorov-Arnold representation theorem, KAN achieves superior function approximation with fewer parameters than traditiona

deep learning

Koopa

Koopa is a deep learning model for time-series forecasting introduced by Yong Liu, Chang Li, Jianmin Wang, and Mingsheng Long at NeurIPS 2023. It addresses the challenge of non-stationarity by disentangling time series into stationary and non-stationary components, then modeling the non-stationary dynamics using a lear

machine learning

Label Propagation

Label Propagation is a graph-based semi-supervised learning algorithm introduced by Zhu and Ghahramani in 2002 that spreads class labels from a small set of labeled nodes to a large set of unlabeled nodes by iteratively diffusing label information along the edges of a similarity graph, exploiting the manifold structure

text mining

Language Identification

Language identification is a natural-language-processing task that automatically detects which language a piece of text is written in. Building on off-the-shelf tools such as langid.py (Lui & Baldwin, 2012) and the efficient classifiers of Joulin et al. (2017), it is widely used to preprocess and filter multilingual da

fluid dynamics

Large Eddy Simulation

Large Eddy Simulation (LES) is a turbulence modeling technique that explicitly resolves large-scale turbulent eddies while modeling small-scale subgrid-scale (SGS) motions. Introduced by Joseph Smagorinsky in 1963, LES represents a middle ground between Reynolds-Averaged Navier-Stokes (RANS) and Direct Numerical Simula

deep learning

Latent Diffusion Models

Latent Diffusion Models (LDMs) are a generative approach introduced by Rombach et al. in 2022 that performs the diffusion process in a compressed latent space rather than pixel space, enabling efficient high-resolution image synthesis. By compressing images into a low-dimensional latent representation using a variation

machine learning

Latent Dirichlet Allocation

Latent Dirichlet Allocation (LDA) is a generative probabilistic model for collections of discrete data, introduced by Blei, Ng, and Jordan in 2003. It treats each document as a mixture of latent topics and each topic as a probability distribution over words, enabling unsupervised discovery of thematic structure across

simulation

Latin Hypercube Sampling

Latin Hypercube Sampling (LHS) is a stratified space-filling design for computer experiments, introduced by McKay, Beckman, and Conover in 1979. It divides each input variable's range into equally probable strata and draws exactly one sample per stratum, ensuring that the full input space is covered with far fewer mode