Enterprise Technology Glossary

6,173 results · page 156 of 309

• Logistic Regression

  Logistic regression is a statistical and machine learning technique for modeling the probability of a categorical, often binary, outcome from input variables, used in enterprises for interpretable risk scoring, classification, and probability estimation within analytics, decision-support, and governance frameworks.

• Logistics Resilience Model

  Logistics Resilience Model is a structured analytical or quantitative framework used to evaluate how logistics and supply chain networks maintain or recover performance under disruptions, supporting enterprise risk management, continuity planning, and design of more robust, adaptable logistics configurations.

• Logistics Route Optimization

  Logistics route optimization is the computational process of designing cost-efficient, constraint-aware transportation routes across a logistics network, used by enterprises to reduce transport cost, meet delivery commitments, support regulatory compliance, and provide route plans to fleets and carrier partners.

• Log Management

  Log management is the process and tooling that collect, centralize, store, and retain log data from systems, networks, and applications so enterprises can support monitoring, incident response, troubleshooting, and regulatory or audit requirements across heterogeneous IT and security environments.

• Log Stream

  Log stream is a continuous, ordered flow of log records produced by systems and applications and processed in near real time, used in enterprises for monitoring, security analytics, observability, and compliance-focused logging across infrastructure and software environments.

• Log-Structured Merge Tree

  Log-structured merge tree (LSM tree) is a write-optimized data structure used in many database storage engines to handle high-throughput inserts and updates, organizing data across memory and disk with sequential writes and compaction for enterprise-scale workloads.

• Long Haul

  Long haul is a large language model architecture and deployment pattern for handling extended, multi-step, and stateful interactions over long contexts in enterprise environments, enabling complex knowledge work while requiring careful design for security, governance, and operational management.

• Long Short-Term Memory

  Long Short-Term Memory is a recurrent neural network architecture that uses memory cells and gating mechanisms to model long-range dependencies in sequence data, which enterprises apply to forecasting, anomaly scoring, and other temporal or ordered data analysis workloads.

• Long Term Evolution

  Long Term Evolution is a 3GPP-standardized mobile broadband technology that uses an all-IP, packet-switched architecture to deliver higher data throughput and lower latency than earlier cellular generations, which matters for enterprise mobility, IoT connectivity, and WAN transport planning.

• Loss Convergence Curve

  Loss convergence curve is a visualization of a machine learning model’s loss value over training iterations or epochs, used by enterprise teams to assess optimizer behavior, detect training issues, and make decisions about training duration, configuration, and resource use.

• Loss Function

  Loss function is a mathematical construct used in machine learning and statistical modeling to measure the difference between predicted and actual values, guiding optimization during training and aligning model behavior with enterprise performance objectives and risk or cost preferences.

• Lossless Ethernet Fabric

  Lossless Ethernet fabric is a data center network architecture that uses Ethernet extensions for congestion management and flow control to avoid frame loss for selected traffic classes, supporting converged storage, RDMA, and latency-sensitive workloads on an Ethernet-based infrastructure.

• Low Code

  Low-code is a class of application development platforms that use visual modeling, configuration, and reusable components to create software with minimal hand-coding, relevant for enterprises that need to deliver custom applications under governance, integration, and security requirements.

• Low-Code Application Platform

  Low-code application platform is a software environment that supports application development and deployment through visual modeling and configuration with minimal hand-coding, while still allowing extensibility through standard programming and integration, which enterprises use to build and govern custom business applications at scale.

• Low Earth Orbit

  Low Earth orbit is an Earth-centered orbital band from roughly 160 to 2,000 kilometers in altitude that hosts many communications, Earth observation, and crewed spacecraft, and it matters to enterprises as an infrastructure layer for connectivity and space-derived data services.

• Low Earth Orbit (LEO) Constellation

  Low Earth orbit constellation is a coordinated network of satellites operating at low Earth altitudes to provide continuous or near-continuous coverage for communications, sensing, and data services, relevant to enterprises that require connectivity, redundancy, and geospatial data beyond terrestrial network reach.

• Low Earth Orbit (LEO) Defense Network

  Low Earth Orbit (LEO) defense network is an integrated military or governmental system of LEO satellites and ground infrastructure that delivers sensing, communications, and missile defense functions, and matters for enterprises involved in secure space, defense, and mission-data architectures.

• Low-Latency Edge Fabric

  Low-latency edge fabric is a distributed networking and compute architecture that connects edge nodes and services to process data close to its source with tightly controlled latency, supporting real-time enterprise workloads while integrating with broader edge-to-cloud and zero trust architectures.

• Low Latency Fabric

  Low latency fabric is a data center or cluster interconnect architecture that minimizes communication delay between servers, storage, and accelerators, enabling efficient execution of parallel, latency-sensitive enterprise workloads such as high-performance computing, artificial intelligence, and real-time analytics.

• Low-Latency Inference Path

  Low-latency inference path is the end-to-end execution route that processes AI or machine learning model requests within strict time bounds, enabling timely responses for interactive, real-time, or near-real-time enterprise applications where bounded response times and predictable performance are required.