Glossary

Master the backpropagation algorithm: how neural networks learn through gradient descent, chain rule, and error propagation. Complete guide with examples and implementations.

Learn how to implement and use the Bag-of-Words model in Natural Language Processing using Python. Discover practical applications in text classification, sentiment analysis, and topic modeling, with code examples using scikit-learn and TF-IDF techniques.

A comprehensive guide to binary and multi-class classification in machine learning. Learn about sigmoid and softmax functions, essential evaluation metrics like accuracy, precision, recall, and F1 score, and practical considerations for model performance. Discover how to handle challenges such as class imbalance, feature engineering, and model selection. Perfect for data scientists and ML practitioners looking to master classification techniques for real-world applications.

A comprehensive guide to clustering in machine learning and data analysis. Explore essential techniques including K-means, hierarchical, density-based, and model-based clustering algorithms. Learn about similarity measures, real-world applications in market segmentation, anomaly detection, and bioinformatics.

Explore Convolutional Neural Networks (CNNs) and how they reshaped computer vision. Learn about network architecture, including convolutional, pooling, and fully connected layers, image recognition processes, and real-world applications in object detection, semantic segmentation, and medical imaging.

Learn data augmentation techniques to expand datasets artificially. Discover image, audio, text, and time series augmentation methods using GANs, transformations, and synthetic data generation.

Master data cleaning techniques for accurate analysis. Learn data preprocessing methods, outlier detection, missing value imputation, and machine learning approaches to ensure data quality and reliability.

An in-depth guide to decision forests in machine learning, covering random forests and gradient boosted trees. Learn about tree construction, ensemble methods, hyperparameter tuning, and real-world applications in finance, healthcare, and marketing. Perfect for data scientists and ML engineers working with tabular data and seeking interpretable, high-performance models.

Explore dimensionality reduction techniques including PCA, t-SNE, and autoencoders. Learn how to combat the curse of dimensionality, improve model performance, and efficiently handle high-dimensional data.

Learn feature engineering techniques for machine learning success. Discover data preprocessing methods, dimensionality reduction, automated feature selection, and real-world applications to boost model performance.

Master feature scaling and normalization techniques for ML. Learn min-max scaling vs standardization with real-world applications and implementation tips.

Learn essential feature selection techniques for machine learning. Discover filter, wrapper, and embedded methods to improve model performance and reduce overfitting.

Explore Generative Adversarial Networks (GANs), their architecture, training process, and applications. Learn about generator and discriminator networks, loss functions, and recent advancements in image synthesis, data augmentation, and text generation. Essential reading for AI researchers, machine learning practitioners, and deep learning enthusiasts interested in generative models.

Explore how boosting algorithms like AdaBoost and Gradient Boosting transform weak learners into powerful predictive models. Discover practical applications in fraud detection, medical diagnosis, and credit risk assessment, with insights on implementation and best practices.

Explore Graph Neural Networks (GNNs) and their applications in analyzing graph-structured data. Learn about key architectures including GCNs, GATs, and GINs, message passing mechanisms, and applications in computer vision, drug discovery, and physics.

Deploying an embedding model on AWS with SIE requires three steps: configure your GPU cluster using the SIE Terraform module, deploy the inference server with Helm, and connect using the SIE SDK. The process takes under 30 minutes and produces a production-grade inference cluster in your own AWS account with automat...

Deploying an embedding model on GCP with SIE requires configuring a GPU cluster using the SIE Terraform module for Google Cloud, deploying the inference server with Helm, and connecting via the SIE SDK. Your data stays within your GCP project, costs are up to 50× lower than managed API pricing, and the full deployme...

Chroma is an open-source vector database designed for simplicity. It runs in-process with Python, requires no infrastructure setup for development, and persists to disk or runs as a server for production. It works with embedding models by accepting vectors at insert time, generated by SIE, and enables cosine simila...

Haystack is an open-source framework for building NLP and RAG pipelines by connecting modular components (document stores, retrievers, readers, and generators) into composable directed acyclic graphs. It works with embedding models by providing retriever components that call the embedding model at query and index ...

Qdrant is an open-source vector database that stores embedding vectors alongside payload metadata and enables fast approximate nearest neighbour (ANN) search, filtered search, and hybrid (dense + sparse) search. It works with embedding models by receiving the vectors they produce (generated by SIE) and indexing th...

SIE (Superlinked Inference Engine) and Infinity are both open-source servers for self-hosting text embedding and reranking models. Infinity is a lightweight, fast single-model server with a focus on OpenAI-compatible API endpoints. SIE is a broader inference platform with multi-model support, LoRA hot-loading, GPU c...

Weaviate is an open-source vector database that stores objects with vector representations and enables semantic search, hybrid search, and filtered retrieval via a GraphQL or REST API. It works with embedding models by accepting vectors at insert time (generated by SIE) and indexing them in HNSW graphs for fast AN...

Master loss functions in machine learning: MSE, cross-entropy, MAE & more. Complete guide to choosing the right loss function for neural network training and model optimization.

Explore neural networks from fundamentals to advanced concepts, including network architecture, backpropagation, and multi-class classification techniques. Learn about activation functions, training processes, and the motivation behind deep learning. Essential reading for AI researchers, machine learning engineers, and developers working with deep neural networks.

Complete guide to machine learning optimizers: SGD, Adam, RMSprop & gradient descent. Learn neural network training algorithms with real-world examples.

A guide to recommendation systems, covering content-based filtering, collaborative filtering, and hybrid approaches. Learn about embeddings, candidate generation techniques, and the role of neural networks in personalization. Essential reading for data scientists, ML engineers, and developers building modern recommendation engines and personalized user experiences.

Master the fundamentals of linear and logistic regression in machine learning. Explore key concepts including features, weights, bias, and optimization techniques. Learn about cost functions, regularization methods, evaluation metrics, and practical considerations for model performance. Essential reading for data scientists and analysts working with predictive modeling and statistical analysis.

Explore Recurrent Neural Networks (RNNs) and their role in processing sequential data. Learn about LSTM architecture, variants like GRUs and Bidirectional LSTMs, and applications in NLP, speech recognition, and time series analysis.

SIE (Superlinked Inference Engine) and TEI (Text Embeddings Inference by Hugging Face) are both open-source servers for self-hosting text embedding models. TEI is a lightweight, single-model server focused on embeddings. SIE is a broader inference platform supporting multiple simultaneous models, rerankers, extracti...

A detailed exploration of Support Vector Machines (SVMs), covering their mathematical principles, types, and real-world applications. Learn about linear and nonlinear classifications, kernel functions, and how SVMs compare to other machine learning algorithms in text classification, image recognition, and financial prediction tasks.

A comprehensive guide to Transformer neural networks, exploring the architecture that reshaped natural language processing. Learn about self-attention mechanisms, encoder-decoder structures, and how Transformers overcome traditional RNN limitations. Discover their applications in language modeling, machine translation, and emerging limitations in computational complexity and interpretability.

A chunking strategy is the approach used to split source documents into smaller segments before encoding them into vectors for a RAG (Retrieval-Augmented Generation) pipeline. The size, overlap, and boundary logic of chunks directly affects retrieval quality. Chunks that are too large compress too much information ...

A document extraction pipeline is an automated system that ingests raw documents (PDFs, Word files, scanned images, HTML pages), extracts structured information from them (text, tables, entities, metadata), and prepares the output for downstream use in search indexes, databases, or RAG systems. It combines OCR, layo...

A LoRA (Low-Rank Adaptation) adapter is a lightweight set of trainable weight matrices added to specific layers of a pre-trained neural network. During fine-tuning, only the LoRA weights are updated; the base model weights remain frozen. This reduces the number of trainable parameters by 100-1000x compared to full ...

A reranker is a model that takes a query and a set of candidate results from an initial retrieval step, and re-scores them for relevance. Unlike embedding models that encode query and documents independently, rerankers compare them jointly, producing more accurate relevance scores at the cost of higher latency.

A reranking pipeline is a two-stage retrieval architecture where a fast first-stage retriever (embedding model + vector DB) fetches a broad set of candidates, and a slower but more accurate second-stage reranker (cross-encoder) re-scores and reorders them. The result is retrieval that combines the scalability of ANN...

A text embedding model is a neural network that converts text (a word, sentence, paragraph, or document) into a dense numerical vector that captures its semantic meaning. Texts with similar meanings produce vectors that are close together in the embedding space, enabling similarity search, clustering, classificati...

A vector database is a database purpose-built for storing, indexing, and querying high-dimensional numerical vectors. Unlike traditional databases that query by exact value or keyword, vector databases find the nearest vectors to a query vector using approximate nearest neighbour (ANN) algorithms, enabling semantic...

A vector index is a data structure that organises high-dimensional vectors to enable fast approximate nearest neighbour (ANN) search. Instead of comparing a query vector to every stored vector, the index groups or graphs vectors in ways that allow search to skip irrelevant regions of the vector space, reducing quer...

An OCR (Optical Character Recognition) model is a machine learning system that extracts text from images, scanned documents, or PDFs. Modern OCR models use deep learning (typically a CNN encoder to detect text regions and a sequence model to decode characters) and can handle handwriting, complex layouts, tables, a...

Approximate Nearest Neighbour (ANN) search finds vectors that are close to a query vector in high-dimensional space (not necessarily the exact closest, but close enough) in milliseconds rather than the seconds or minutes exact search would require. It is the core retrieval algorithm inside every vector database an...

BGE-M3 is an open-source text embedding model developed by BAAI (Beijing Academy of Artificial Intelligence) that supports three retrieval modes simultaneously: dense retrieval, sparse retrieval, and multi-vector (ColBERT-style) retrieval. It supports 100+ languages and is one of the highest-performing general-purpo...

ColBERT (Contextualized Late Interaction over BERT) is a neural retrieval model that encodes queries and documents into token-level vector representations and scores their relevance using a MaxSim operation that finds the best-matching document token for each query token, then sums those scores. It achieves near c...

GPU utilisation in inference refers to the percentage of a GPU's compute capacity actively used when serving model predictions. High GPU utilisation (70-95%) means you're getting maximum throughput per dollar. Low utilisation (under 30%) means you're paying for idle hardware. Efficient batching, model loading strate...

Hybrid search combines dense vector (semantic) search with sparse keyword (BM25-style) search, merging results from both to produce a single ranked list. It captures the strengths of each approach (semantic understanding from dense retrieval and exact-term precision from sparse retrieval), making it more accurate t...

Instruction-following embedding is a technique where a natural language instruction is prepended to the input text before encoding, telling the embedding model what type of representation to produce. Instead of always generating the same vector for a given text, the model adapts its output based on the task describ...

Late interaction retrieval is a neural search architecture where the query and document are encoded independently into token-level vectors, and their similarity is computed at retrieval time using a fine-grained token matching function (typically MaxSim). It sits between bi-encoders (fast but less accurate) and cros...

Multi-vector search is a retrieval technique where each document is represented by multiple vectors (one per token or passage) rather than a single fixed-size vector. At query time, the query's token vectors are compared against all document token vectors, enabling fine-grained token-level matching that captures n...

Retrieval-Augmented Generation (RAG) is an AI architecture that combines a retrieval system with a large language model (LLM). Instead of relying solely on the LLM's training data, RAG first retrieves relevant documents from a knowledge base and passes them as context to the LLM, producing answers that are grounded...

Self-hosted inference is the practice of running AI model inference on your own infrastructure (your own cloud account such as AWS, GCP, Azure, or on-premises hardware) rather than sending requests to a third-party managed API. You control the hardware, the models, the configuration, and crucially, where your data goes.

Semantic search is a retrieval technique that finds results based on the meaning of a query rather than exact keyword matches. Instead of matching words character-by-character, it converts text into vector embeddings and retrieves items whose embeddings are closest to the query's embedding in high-dimensional space.