Machine Learning

AI models do not have to be huge, slow or expensive. Distilling Intelligence explores how to build smaller models that perform at scale, then secure the APIs that serve them. From compression and distributed serving to extraction attacks, observability and incident response, this book covers what it takes to run AI reliably in the real world.

Aprende a combinar SQL, Python, estadística y Machine Learning para priorizar posibles fraudes en retail. Un enfoque práctico, técnico y aplicado, con ejemplos, visualizaciones y códigos de apoyo.

Build real machine learning applications with TensorFlow.js, from your first model to production-ready systems. Learn how to train and deploy models in the browser and Node.js, tackle real-world challenges and explore modern techniques with practical JavaScript and TypeScript examples you can run and adapt.

This practical guide explains how to calculate, interpret, and improve OEE using real-world examples, enabling engineers, supervisors, and Lean practitioners to increase productivity, reduce downtime, and achieve world-class manufacturing performance.

Building an LLM demo is easy. Running one reliably at scale is not. Production LLMOps shows you how to design, deploy and operate real-world LLM systems, covering RAG, agents, fine-tuning, evaluation, CI/CD, observability, security and more, with practical code you can adapt for production.

A from-scratch AI/ML course that treats you like an engineer, not a tourist — three modules covering language-as-numbers, the math foundations, and neural networks, each concept built in raw Python first, then PyTorch, so you always know what's really happening under the hood.

The eight volumes before this one each teach a single domain, and each is deliberatelyself-contained. The exams do not ask about domains — they ask about systems, and a system spansall of them at once. "A retrieval system is slow, expensive and occasionally wrong — which do you fix first?" needsretrieval, optimisation and operations together. Every one of the 150 questions here crosses atleast two domains, because that is the specific thing a single-domain book cannot teach you.

"The model is safe." "The filter is effective." "We red-teamed it and found nothing." "Wereviewed a sample and it looked fine." Each of those is a claim about evidence, and each is unfalsifiable as stated. Reviewing thirtyoutputs and finding no problems is consistent with a failure rate of one in eleven. Threehundred tests against a thousand possible failure modes leave seven hundred nobody looked at. This book supplies the missing arithmetic — and it was written without generating a singleharmful output to study.

Quantise the weights. Batch the requests. Add a draft model. Prune what you can. Shorten thecache. A long list of techniques and a short list of reasons. This book measures the reason first: the machine streams memory at 22.54 GB/s and computes at116.58 GFLOP/s, so anything performing fewer than 5.17 operations per byte is waiting. Decodingperforms 0.22. Every technique in the book is then measured against that one number — andseveral turn out to be addressing the wrong side of it.

Понятный иллюстрированный путеводитель по ключевым концепциям и практическим применениям больших языковых моделей. Идеально подходит для работы над проектами, подготовки к собеседованиям и самостоятельного обучения.

هذا الكتاب هو دليل مختصر وموضّح بالرسوم لأي شخص يرغب في فهم الآلية الداخلية للنماذج اللغوية الضخمة، سواء في سياق المقابلات أو المشاريع أو بدافع الفضول الشخصي.

The practitioner's guide to policy-space response oracles: one 84-line loop, one readable repository, and every claim backed by a committed experiment — including the ones that came out wrong.

Use LoRA because it forgets less. Keep the learning rate low. Mix in some general data. Stopearly. Four pieces of advice everyone repeats and almost nobody quantifies. This book measures each one — and two turn out to be doing something other than what they arecredited with. LoRA's retention is governed by a scaling constant, not by its parameterefficiency. Early stopping bounds the damage without separating it from the gain. Everyexperiment reports two numbers: the new task, and the capability you were not trying tochange.

Chunk at 500 with 50 of overlap. Retrieve five. Add a reranker. Use hybrid search.Repeated everywhere, justified almost nowhere. This book measures them instead, on corpora built so that the correct answer follows fromhow they were made. Several results contradicted the advice: the hybrid combination lost inevery condition tested, neither reranker produced an effect above the noise floor, andattribution collapsed to 0.170 while retrieval was still reporting 0.920.

Vector databases give you a similarity metric, an index with four tuning parameters, acompression option and a way to filter — and almost no guidance on what any of it costs. This book builds each one from scratch and measures it. The proximity graph, the invertedfile, the product quantiser: implemented in NumPy, included with the text, and checkedagainst exhaustive search — so every recall figure here is exact rather than estimated.