按方法、领域与证据探索科学。

Lattice Boltzmann Method

The Lattice Boltzmann Method (LBM) is a kinetic theory-based computational approach to fluid dynamics that discretizes the Boltzmann equation on a lattice grid. Developed by McNamara and Zanetti in 1988, LBM computes fluid behavior by tracking the distribution of particle velocities at discrete lattice nodes rather tha

quantum computing

Lattice QCD

Lattice Quantum Chromodynamics (LQCD) is a computational method for studying quantum chromodynamics (QCD)—the theory of strong nuclear forces—by discretizing spacetime onto a lattice and simulating quark and gluon dynamics. Introduced by Kenneth Wilson in 1974, LQCD is the only known approach for non-perturbative calcu

cryptography

Lattice-Based Cryptography

Lattice-based cryptography is a class of cryptosystems whose security is derived from the computational hardness of lattice problems, particularly the shortest vector problem (SVP) and learning with errors (LWE). First proposed by Miklós Ajtai in 1996, lattice-based approaches have gained prominence as the leading cand

genetics

LD Block Analysis

Linkage disequilibrium (LD) block analysis is a genomic method that partitions the human genome into distinct haplotype blocks—regions of limited recombination where variants are in strong statistical association. First systematically described by Gabriel and colleagues in 2002, this approach reveals the underlying str

deep learning

LDA Topic Model

Latent Dirichlet Allocation (LDA) is a probabilistic generative model introduced by Blei, Ng, and Jordan in 2003 that discovers hidden thematic structure in large text collections by representing each document as a mixture of latent topics and each topic as a probability distribution over vocabulary words.

telecommunications

LDPC Codes

LDPC codes, invented by Robert Gallager in 1962 and rediscovered in the 1990s by MacKay, are linear error-correcting codes defined by sparse parity-check matrices. They achieve performance within 0.4 dB of the Shannon limit with iterative belief-propagation decoding and have become the standard for modern wireless (WiF

mining engineering

Lerchs-Grossmann Algorithm

The Lerchs-Grossmann Algorithm is a graph-theoretic method for determining the ultimate pit limit in open-pit mining operations. Introduced by Helmut Lerchs and Israel Grossmann in 1965, it maximizes the net present value of extracted ore while respecting slope stability constraints. This algorithm forms the theoretica

fluid dynamics

Level Set Method

The Level Set Method is an implicit interface tracking technique introduced by Osher and Sethian in 1988 for moving boundary problems and multiphase flows. Rather than explicitly tracking the interface, level sets represent it as the zero level set (contour) of a signed distance function φ. This approach elegantly hand

thermodynamics

Levelized Cost of Energy

Levelized Cost of Energy (LCOE) is a standardized metric that spreads the total lifecycle cost of an energy project over its lifetime energy output. It enables fair comparison of electricity generation technologies with different capital structures, operating costs, and lifetimes. LCOE is widely used for technology eva

text mining

Lexical Diversity

Lexical diversity analysis quantifies how varied the vocabulary of a text is — how rich an author's word choice is — using measures such as the type-token ratio (TTR), MTLD, vocd-D, and Yule's K. The MTLD and vocd-D measures were validated by McCarthy and Jarvis (2010), building on earlier work by Tweedie and Baayen (1

text mining

Lexical Substitution

Lexical substitution is a natural-language-processing task — formalised by McCarthy and Navigli through the SemEval shared task series starting in 2007 — that replaces a target word in a sentence with a semantically equivalent alternative that preserves the meaning of the surrounding context. It draws on synonym resour

text mining

Lexicon-Based Sentiment Analysis

Lexicon-based sentiment analysis computes sentiment at the word level using prebuilt sentiment dictionaries such as AFINN (Nielsen, 2011), SentiWordNet, VADER (Hutto & Gilbert, 2014), and the NRC Emotion Lexicon. It scores text by looking words up in a dictionary of charged terms, so it requires no labelled training da

model evaluation

Lift and Gain Chart

Lift and gain charts visualize classifier performance by showing how much better the model performs compared to random selection, particularly useful for ranking or scoring tasks where you select a top percentage of samples. They are widely used in marketing, credit scoring, and fraud detection.

applied physics

Light Curve Analysis

Light curve analysis is the study of the brightness variation of a celestial object over time, used to detect and characterize exoplanets, eclipsing binaries, and variable stars. When a planet transits in front of its host star, the star's brightness dips slightly. By analyzing these photometric signatures, astronomers

machine learning

LightGBM

LightGBM is Microsoft's gradient boosting decision tree implementation, introduced by Ke and colleagues in 2017, that grows trees leaf-wise and bins features into histograms for speed. On large datasets it is much faster than XGBoost while retaining strong predictive accuracy.

meteorology

Lightning Jump

The lightning jump is a rapid increase in lightning activity (number of flashes per unit time) that precedes severe weather including hail, heavy rain, and tornadoes. This phenomenon, observed using satellite or ground-based lightning detection networks, is an operational diagnostic tool for real-time severe weather wa

deep learning

LightTS

LightTS is a lightweight, MLP-based architecture for multivariate time-series forecasting introduced by Tianping Zhang and colleagues in 2022. Motivated by the observation that simpler models can match or surpass heavy Transformer-based architectures, LightTS applies an interval-sampling strategy to decompose long inpu

machine learning

LIME

LIME, introduced by Ribeiro, Singh, and Guestrin in 2016, explains the predictions of any black-box classifier or regressor by building a simple, locally faithful surrogate model around a single prediction of interest. Rather than explaining the global model, LIME focuses on why a specific instance was classified the w

cryptography

Linear Cryptanalysis

Linear cryptanalysis is a known-plaintext attack that exploits linear approximations of a cipher's non-linear transformations to recover secret key bits. Introduced by Mitsuru Matsui in 1993, linear cryptanalysis provides practical attacks on ciphers like DES with computational complexity less than brute force. The tec

machine learning

Linear Discriminant Analysis

Linear Discriminant Analysis is a supervised method for dimensionality reduction and classification, introduced by Ronald A. Fisher in 1936, that finds linear combinations of features which maximally separate predefined classes while preserving as much class-discriminatory information as possible. It simultaneously ser

acoustics

Linear Predictive Coding

Linear Predictive Coding (LPC) is a powerful signal processing technique for modeling and compressing speech by assuming each speech sample can be predicted from a linear combination of previous samples. Pioneered by Burg and Makhoul in the 1970s, LPC is the foundation of speech codecs, speech synthesis, speaker recogn

control theory

Linear Quadratic Gaussian

The Linear Quadratic Gaussian (LQG) controller combines the Linear Quadratic Regulator (LQR) with a Kalman Filter to handle stochastic systems with measurement noise and process noise. Developed by Kalman and later formalized by Athans and others, LQG is the natural stochastic extension of LQR and remains the gold stan

control theory

Linear Quadratic Regulator

The Linear Quadratic Regulator (LQR) is a classical optimal control algorithm that computes a linear feedback law to minimize a quadratic cost function for a linear dynamical system. Introduced by Kalman in 1960, LQR provides a provably optimal, closed-form solution for linear systems and remains fundamental in control

machine learning

Linear Regression (ML)

Linear regression fits a straight-line relationship between one or more input features and a continuous numeric outcome by minimising the sum of squared prediction errors. As a machine-learning model it is trained on labeled examples and evaluated on held-out data, making it the simplest supervised learning baseline fo

text mining

Linguistic Acceptability Assessment

Linguistic acceptability assessment is a natural-language-processing task that automatically estimates whether a sentence would be judged grammatically acceptable by a native speaker of the target language. Grounded in Chomsky's (1957) distinction between grammatical and ungrammatical utterances, the task was formalise

linguistics

Linguistic Ethnography

Linguistic Ethnography is a qualitative research approach combining ethnographic fieldwork with detailed linguistic analysis to understand language use in cultural context. Developed by researchers like Ben Rampton, it examines how people use language within communities, institutions, and social interactions while payi

network analysis

Link Prediction

Link prediction is a network-analysis task that estimates which edges are missing from an observed graph or which edges are likely to form in the future. Formalised by Liben-Nowell and Kleinberg (2003, 2007), it covers a spectrum of approaches — from simple structural similarity indices such as Common Neighbors, Jaccar

electrical engineering

Load Forecasting

Load forecasting predicts future electrical demand on power systems across various time horizons: minutes to hours (short-term), days to weeks (medium-term), and months to years (long-term). Accurate forecasting is essential for economic dispatch, unit commitment, and system reliability. Methods range from classical st

electrical engineering

Load-Pull

Load-Pull is an experimental technique for characterizing and optimizing RF power amplifier performance under varying load and source impedance conditions. Introduced by Davidson et al. in 1990, load-pull measurements vary the load impedance seen by the amplifier while recording output power, efficiency, and linearity.

machine learning

Local Outlier Factor

Local Outlier Factor (LOF) is a density-based, unsupervised anomaly detection algorithm introduced by Breunig, Kriegel, Ng, and Sander in 2000. It assigns each data point a continuous outlier score that quantifies how isolated that point is relative to its local neighborhood, enabling detection of anomalies that global

machine learning

Locally Linear Embedding

Locally linear embedding, introduced by Sam Roweis and Lawrence Saul in 2000, is a manifold-learning method for nonlinear dimensionality reduction. It assumes that although data may curve through a high-dimensional space, each point and its neighbours lie approximately on a flat patch. LLE captures each point as a weig

machine learning

LOESS

LOESS (locally estimated scatterplot smoothing), introduced by William Cleveland in 1979 and extended with Susan Devlin in 1988, fits a smooth curve through data by performing a separate weighted polynomial regression in the neighbourhood of each point. Nearby observations count more than distant ones, so the method fo

thermodynamics

Log Mean Temperature Difference

The Log Mean Temperature Difference (LMTD) method is a fundamental tool for calculating heat transfer rates in heat exchangers. It defines the effective temperature difference between two fluids as the logarithmic average of the temperature differences at the inlet and outlet. This method enables engineers to size and

model evaluation

Log-Loss (Cross-Entropy Loss)

Log-loss measures the difference between predicted probabilities and actual labels, penalizing confident wrong predictions more than uncertain ones. It is a standard loss function in machine learning optimization and evaluates probabilistic classifier calibration.

electrical engineering

Logic Synthesis

Logic Synthesis is the automated conversion of high-level hardware descriptions (RTL in Verilog/VHDL) into optimized gate-level netlists. Pioneered by Brayton et al. at UC Berkeley in the 1980s-1990s, logic synthesis transforms behavioral specifications into physical implementations, optimizing for area, speed, and pow

deep learning

Long Short-Term Memory

Long Short-Term Memory (LSTM) is a gated recurrent neural network architecture introduced by Hochreiter and Schmidhuber in 1997. It was designed to learn dependencies across long sequences by using dedicated memory cells and three learned gates — forget, input, and output — that control what information is retained, up

deep learning

Longformer / BigBird

Long-sequence Transformers such as Longformer (Beltagy, Peters & Cohan, 2020) and BigBird (Zaheer et al., 2020) replace the standard Transformer's O(n²) attention with sparse attention patterns that scale linearly, O(n), with sequence length. This lets a single model attend over thousands of tokens — full documents, le

psychometrics

Longitudinal Item Analysis

Longitudinal item analysis examines how the statistical properties of individual scale items — difficulty, discrimination, factor loadings, and fit — remain stable or change systematically across repeated measurement occasions. It is the item-level foundation of longitudinal measurement validity.

quantitative finance

Longstaff-Schwartz Method

The Longstaff-Schwartz method (2001) is a Monte Carlo algorithm for pricing American options and Bermudan swaptions by approximating the optimal exercise boundary via least-squares regression. It has become the industry standard for pricing path-dependent derivatives where analytical solutions do not exist.

deep learning

LoRA and PEFT

LoRA (Low-Rank Adaptation), introduced by Hu et al. in 2022, and the broader family of parameter-efficient fine-tuning (PEFT) methods adapt large pretrained language models to new tasks by training only a small number of extra parameters instead of every weight in the model. This makes fine-tuning possible with far les

deep learning

LSTM

LSTM (Long Short-Term Memory) is a recurrent neural network architecture, introduced by Sepp Hochreiter and Jürgen Schmidhuber in 1997, that can learn long-term dependencies in sequential data and is widely used for time-series and sequence prediction. It keeps an internal memory that lets information persist across ma

computer vision

Lucas-Kanade Optical Flow

The Lucas-Kanade method, introduced by Bruce Lucas and Takeo Kanade in 1981, is a foundational technique for estimating optical flow—the apparent motion of objects in image sequences. By computing pixel-level motion vectors, the Lucas-Kanade algorithm tracks feature displacements between consecutive frames, enabling ob

thermodynamics

Lumped Capacitance Method

The Lumped Capacitance Method is a simplification technique for solving unsteady-state heat transfer problems. It assumes that thermal properties are uniform throughout a solid body and that temperature variations within the object are negligible. This approach enables engineers to solve complex transient heat conducti

bioinformatics

Machine learning-assisted ChIP-seq peak calling

Machine learning-assisted ChIP-seq peak calling extends classical statistical peak detection with supervised or unsupervised learning models that distinguish genuine protein-binding sites from background noise. By training on sequence composition, read coverage profiles, and epigenomic features, these methods improve s

bioinformatics

Machine learning-assisted copy number variation analysis

Machine learning-assisted CNV analysis applies supervised, unsupervised, or deep learning algorithms to detect genomic regions that are duplicated or deleted relative to a reference genome. Rather than relying on fixed statistical thresholds, ML models learn discriminative patterns from read-depth signals, allele frequ

bioinformatics

Machine learning-assisted epigenome-wide association study

Machine learning-assisted EWAS integrates conventional epigenome-wide association testing with machine learning models to identify DNA methylation sites associated with a phenotype of interest. By combining the statistical rigour of EWAS with the pattern-recognition power of algorithms such as elastic net, random fores

bioinformatics

Machine learning-assisted expression quantitative trait loci analysis

Machine learning-assisted eQTL analysis integrates supervised learning models — ranging from elastic-net regression to deep neural networks — into the classical eQTL framework to predict and map genetic variants that regulate gene expression. By training predictive models on reference panels (e.g., GTEx), the approach

bioinformatics

Machine learning-assisted genome-wide association study

Machine learning-assisted GWAS integrates classical genome-wide association testing with machine learning models to improve the detection of genetic variants associated with complex traits. Where traditional GWAS tests each single nucleotide polymorphism (SNP) independently using linear or logistic regression, ML-GWAS

bioinformatics

Machine learning-assisted metabolomics analysis

Machine learning-assisted metabolomics analysis is an integrative bioinformatics pipeline that couples untargeted or targeted metabolite profiling — via mass spectrometry or NMR — with supervised and unsupervised ML algorithms to discover biomarkers, classify phenotypes, and model metabolic states. By handling the extr

bioinformatics

Machine learning-assisted microbiome diversity analysis

Machine learning-assisted microbiome diversity analysis integrates classical alpha and beta diversity metrics with supervised or unsupervised ML models to classify host phenotypes, identify discriminant taxa, and uncover community-level signatures from 16S rRNA or shotgun metagenomic data. It extends traditional divers

bioinformatics

Machine learning-assisted pathway enrichment analysis

Machine learning-assisted pathway enrichment analysis integrates classical statistical pathway enrichment methods — such as over-representation analysis or gene set enrichment analysis — with machine learning algorithms to improve sensitivity, handle high-dimensional omics data, and uncover non-linear biological patter

bioinformatics

Machine learning-assisted phylogenetic analysis

Machine learning-assisted phylogenetic analysis integrates supervised, unsupervised, or deep learning models into the evolutionary tree inference workflow to improve speed, accuracy, or scalability beyond what classical maximum-likelihood and Bayesian methods achieve alone. Applications range from substitution model se

bioinformatics

Machine learning-assisted RNA-seq differential expression

Machine learning-assisted RNA-seq differential expression analysis augments classical statistical DE testing (DESeq2, edgeR, limma-voom) with ML models — including neural networks, random forests, and variational autoencoders — to better handle the high dimensionality, zero-inflation, and batch effects inherent in RNA-

bioinformatics

Machine learning-assisted sequence alignment

Machine learning-assisted sequence alignment uses statistical learning models — including deep neural networks and protein language models — to compute biologically meaningful alignments between nucleotide or amino acid sequences. By learning substitution patterns and structural constraints from large training corpora,

bioinformatics

Machine learning-assisted single-cell RNA-seq analysis

Machine learning-assisted single-cell RNA sequencing (scRNA-seq) analysis integrates supervised, unsupervised, and deep generative models into the standard scRNA-seq workflow to handle the unique challenges of single-cell data: extreme sparsity, high dimensionality, technical noise, and batch effects across experiments

bioinformatics

Machine learning-assisted variant calling

Machine learning-assisted variant calling uses statistical learning models — most notably convolutional neural networks — to distinguish genuine genomic variants (SNPs, indels) from sequencing artifacts in aligned short- or long-read data. Unlike heuristic callers that rely on hand-crafted filters, ML-based approaches

text mining

Machine Reading Comprehension

Machine reading comprehension (MRC), popularised by the SQuAD benchmark of Rajpurkar, Zhang, Lopyrev and Liang (2016), is a natural-language-processing task in which a model reads a given passage and answers multiple-choice or open-ended questions about it. It turns a passage plus a question into a machine-generated an

text mining

Machine Translation

Machine translation (MT) is a natural-language-processing task that automatically converts text in one language into another. Modern MT is built on neural sequence-to-sequence models — the attention mechanism introduced by Bahdanau et al. (2015) and the transformer architecture of Vaswani et al. (2017) — and it widens

model evaluation

Macro-averaged F1

Macro-averaged F1 computes the F1-score independently for each class and then takes the unweighted arithmetic mean. It treats all classes equally, regardless of their frequency in the dataset, making it useful for imbalanced multi-class problems.

aerospace

Madgwick Filter

The Madgwick Filter is a computationally lightweight attitude estimation algorithm that fuses inertial measurements (accelerometer, gyroscope) with magnetic measurements (magnetometer) to compute a quaternion orientation. Introduced by Sebastian Madgwick in 2010, the algorithm uses gradient descent optimization to mini