29,258 technical terms and definitions
data augmentation, strategies advanced techniques, cifar passed, data augmentation strategies & advanced techniques (mixup cutmix), data augmentation strategies, (mixup cutmix)
Data Augmentation: Strategies & Advanced Techniques (Mixup, CutMix) Introduction & Motivation Data augmentation: artificially expand dataset. Geometric: rotation, flipping…
Data Augmentation Strategies Introduction & Motivation Data Augmentation Strategies addresses a central problem in vision processing and augmentation-diversity prediction: how…
data augmentation and regularization, data augmentation regularization, data augmentation, augmentation regularization
Data Augmentation and Regularization Introduction & Motivation Data Augmentation and Regularization addresses a central problem in ml-training process control…
augmentation strategy deep learning, mixup cutmix augmentation, randaugment autoaugment, image augmentation technique
Data Augmentation is the training technique that artificially expands and diversifies the training dataset by applying label-preserving transformations to existing examples…
cutout cutmix mixup augmentation, autoaugment policy, augmentation invariance, test time augmentation
Data Augmentation Techniques is the family of methods that artificially expand training data diversity through geometric transformations, color perturbations, and mixing…
data cache optimization techniques, data cache optimization, cache optimization techniques, data cache, cache optimization, optimization techniques
Data Cache Optimization Techniques Introduction & Motivation Data Cache Optimization Techniques addresses a central problem in performance-optimization process control…
documentation
Data Card is the standardized documentation framework that provides comprehensive metadata about datasets used in machine learning — describing data collection methods…
evaluation
Data Card is a documentation artifact that records dataset provenance, collection methods, labeling process, and ethical considerations - It is a core method in modern AI…
AI data center, PUE, power usage effectiveness, liquid cooling, rack density, Open Compute Project, warehouse-scale computer, InfiniBand
A data center is a facility that houses large numbers of networked computers together with the power, cooling, networking and physical security needed to keep them running…
data center cooling efficiency drift prediction, data center cooling, center cooling efficiency, cooling efficiency drift, efficiency drift prediction, data center
Data Center Cooling Efficiency Drift Prediction with Machine Learning Introduction & Motivation Data Center Cooling Efficiency Drift Prediction with Machine Learning addresses…
data center processor power efficiency prediction, data center processor, center processor power, processor power efficiency, power efficiency prediction, data center
Data Center Processor Power Efficiency Prediction with Machine Learning Introduction & Motivation Data Center Processor Power Efficiency Prediction with Machine Learning addresses…
data center switch fabric latency prediction, data center switch, center switch fabric, switch fabric latency, fabric latency prediction, data center, center switch
Data Center Switch Fabric Latency Prediction with Machine Learning Introduction & Motivation Data Center Switch Fabric Latency Prediction with Machine Learning addresses a central…
data quality, data labeling, data augmentation advanced, data flywheel
Data-Centric AI is the paradigm that prioritizes systematic improvement of training data quality, diversity, and labeling consistency over model architecture changes — recognizing…
data cleansing, deduplication, fuzzy matching, validation rules, missing value handling
Data cleaning is the process of detecting and correcting errors, inconsistencies, and gaps in a dataset before it's used for analysis or modeling — fixing malformed entries…
code ai
Data Clumps are a code smell where the same group of 3 or more data items repeatedly appear together across function parameter lists, class fields, and object initializations…
automation
Data collection automatically gathers process data and metrology results via automation systems, enabling SPC, traceability, and advanced analytics. Data types: (1) Summary…
Column-Wise and Row-Wise Completeness Profiling Introduction & Motivation Column-Wise and Row-Wise Completeness Profiling addresses a central problem in (subject) assurance…
training
Data-constrained regime is the training regime where model performance is primarily limited by insufficient effective data rather than compute or model size - it indicates that…
evaluation
Data contamination occurs when test data appears in training data, artificially inflating benchmark scores. The problem: Model memorizes test examples rather than learning…
evaluation
Data contamination detection is the process of checking whether evaluation benchmark data has been inadvertently included in a model's training set. When test data leaks into…
data quality, quality over quantity, Textbooks Are All You Need, phi-1, Gunasekar 2023, FineWeb, FineWeb-Edu, data filtering, deduplication
Data curation is the work of selecting, cleaning, filtering, deduplicating, balancing, and sometimes rewriting or generating the examples a model is trained on, so that…
data quality
Data deduplication is the process of identifying and removing duplicate or near-duplicate examples from a dataset. It is a critical data quality step for training language models…
data quality
Data deduplication is the process of identifying and removing repeated or near-repeated content from training corpora - it improves data efficiency, reduces memorization risk…
data deduplication storage, data deduplication, deduplication storage
Data Deduplication Storage Introduction & Motivation Data Deduplication Storage addresses a central problem in systems process control and deduplication-overhead-latency…
optimization
Data-Dependent Initialization is a weight initialization approach that uses a batch of real training data to calibrate initial weights — adjusting weight magnitudes and biases…
Prior Distribution Matching for Truncated Datasets Introduction & Motivation Prior Distribution Matching for Truncated Datasets addresses a central problem in (subject) assurance…
mlops
Data drift (also called dataset shift or distribution shift) occurs when the statistical properties of the input data that a deployed model receives in production differ from…
data driven process innovation, data driven, driven process, ct th wip, data-driven process, process innovation
Data-Driven Process Innovation Introduction & Motivation Data-Driven Process Innovation addresses a central problem in process discovery and operational improvement: how to turn…
computer vision
Data efficiency of ViT measures the ability of transformer vision models to reach strong accuracy with limited labeled examples - this efficiency depends heavily on architectural…
data encryption at rest and transit, data encryption rest, encryption rest transit, data encryption, encryption rest, rest transit
Data Encryption at Rest and Transit Introduction & Motivation Data Encryption at Rest and Transit addresses a central problem in ml-security process control…
data engineer, data platform, etl pipeline, elt pipeline, data architecture
Data engineering definition and system boundary. Data engineering designs, builds, and operates the systems that collect, validate, transform, store, govern, and serve data…
Shannon Entropy and Information Density Data Audit Introduction & Motivation Shannon Entropy and Information Density Data Audit addresses a central problem in (subject) assurance…
parsing, scraping
Data Extraction with LLMs Unstructured to Structured Extraction LLMs excel at extracting structured data from unstructured text, emails, documents, and web pages. Basic Extraction…
data quality
Data filtering is the process of systematically removing low-quality, irrelevant, harmful, or redundant examples from a training dataset to improve model performance. In the era…
data quality
Data filtering strategies is multi-stage methods for screening and selecting high-value training samples from raw corpora - It combines source rules, statistical signals…
pandas DataFrame, Wes McKinney, vectorization, Dask, Apache Spark DataFrame, Polars
A data frame is a tabular, in-memory data structure — rows and labeled columns, each column holding a single data type — that serves as the primary working format for data…
model compression
Data-Free Distillation is a knowledge distillation technique that works without access to the original training data — using the teacher model itself to generate synthetic…
Data-Free Knowledge Distillation Introduction & Motivation Data-Free Knowledge Distillation addresses a central problem in reducing the size and cost of compression models: how…
Data-Free Knowledge Distillation (Extended Study) Introduction & Motivation Data-Free Knowledge Distillation (Extended Study) addresses a central problem in reducing the size…
Data-Free Knowledge Distillation (Extended Study) Introduction & Motivation Data-Free Knowledge Distillation (Extended Study) addresses a central problem in reducing the size…
Data-Free Knowledge Distillation (Extended Study) Introduction & Motivation Data-Free Knowledge Distillation (Extended Study) addresses a central problem in reducing the size…
Data Freshness and Staleness Assessment Introduction & Motivation Data Freshness and Staleness Assessment addresses a central problem in (subject) assurance and (target)…
Data Governance and Compliance Checking Introduction & Motivation Data Governance and Compliance Checking addresses a central problem in (subject) assurance and (target)…
data governance policy enforcement, data governance policy, governance policy enforcement, data governance, governance policy, policy enforcement
Data Governance Policy Enforcement Introduction & Motivation Data Governance Policy Enforcement addresses a central problem in ml-ops process control…
Data Imputation Strategy Selection Introduction & Motivation Data Imputation Strategy Selection addresses a central problem in (subject) assurance and (target) management: how…
entity resolution, schema mapping, master data management MDM, Fivetran Airbyte
Data integration is the process of combining data from multiple separate sources — different databases, APIs, file systems, departments, or acquired companies — into a single…
Data Integration Quality Assurance Introduction & Motivation Data Integration Quality Assurance addresses a central problem in (subject) assurance and (target) management: how…
data annotation, human annotation, model assisted labeling, ground truth, preference labeling
Data labeling turns raw examples into the supervised signal a model trains on, and the single hardest thing to internalize about it is that a label is a measurement, not ground…
cloud data lake, object storage data lake, lakehouse, data lake architecture
Data lake definition and system boundary. A data lake is a centralized storage architecture for structured, semi-structured, and unstructured data in durable scalable object…
data lakehouse architecture, data lakehouse, lakehouse architecture
Data Lakehouse Architecture Introduction & Motivation Data Lakehouse Architecture addresses a central problem in systems process control and schema-on-read-flexibility prediction…