Machine learning workflow
A machine learning workflow is an organized sequence of activities for formulating, developing, evaluating, deploying, operating, and maintaining a machine learning system.
Distributed Machine Learning Patterns
The stages and activities involved in developing, deploying, operating, maintaining, and retiring machine learning models and systems.
- How it is used
- Connects ingestion, training, and serving components into reusable end-to-end machine learning systems.
- How deep it goes
- Defines pipelines and dependency forms, then elaborates composition, concurrency, memoization, and failure-handling trade-offs.
- Examples
- Uses YouTube-8M entity tagging and a Fashion-MNIST image-classification pipeline with multiple models, selection, serving, and aggregated predictions.
- What you can do after
- Apply it via Argo Workflows to compose and optimize an end-to-end pipeline with connected component steps and memoization.
Read how Distributed Machine Learning Patterns explains machine learning workflow →
Machine Learning Platform Engineering
The end-to-end sequence of activities through which a machine learning solution progresses from problem formulation and experimentation through deployment, operation, and retirement.
- How it is used
- Organizes ML projects from problem definition through production operation and retirement, with room to revisit earlier work.
- How deep it goes
- Explains phases, iteration, and automation through production scenarios; leaves detailed implementation of individual tools to their respective treatments.
- Examples
- Uses a changing used-car price predictor, a training setback that prompts data rework, and an income-classification deployment.
- What you can do after
- Practice/perform an iterative workflow that aligns a problem, automates repeatable stages, and responds to production changes.
Read how Machine Learning Platform Engineering explains machine learning workflow →
Deep Learning with Python, Third Edition
The end-to-end sequence of steps for defining, building, evaluating, deploying, and maintaining a machine learning system.
- How it is used
- Frames the full practical path for solving machine learning problems end to end.
- How deep it goes
- Describes the major stages and key checks, but stops short of detailed deployment mechanics or advanced theory.
- Examples
- Uses movie reviews, news wires, house prices, spam, recommendations, fraud, click-through rate, anomalies, and satellite-image tasks.
- What you can do after
- Operate a machine learning project from task framing through monitoring and the next model generation.
Read how Deep Learning with Python, Third Edition explains machine learning workflow →
Graph Algorithms for Data Science
A machine learning workflow is a sequence of processing and modeling steps for training or applying machine learning models.
- How it is used
- Organizes graph data preparation, feature extraction, and model training around prediction or classification tasks.
- How deep it goes
- Explains preprocessing rationale and feature construction at a high level, deferring detailed graph transformations to later techniques.
- Examples
- Uses multi-type graph reshaping, missing-value prediction, future-behavior classification, community detection, and centrality as examples.
- What you can do after
- Apply it via a graph database query language to prepare algorithm inputs and construct features for machine learning.
Read how Graph Algorithms for Data Science explains machine learning workflow →
Deep Learning with R, Third Edition
A machine learning workflow is an organized sequence of activities for defining a problem, preparing data, developing a model, and deploying and maintaining the resulting system.
- How it is used
- Organizes machine learning work from problem framing through production monitoring instead of treating model training as the entire task.
- How deep it goes
- Provides a procedural blueprint covering data, validation, baselines, capacity, regularization, deployment, and monitoring without detailed algorithmic derivations.
- Examples
- Uses satellite images for archaeological-site prediction and contrasts real-world work with ready-made tutorial datasets.
- What you can do after
- Apply the workflow to frame a problem, evaluate models safely, improve generalization, and plan deployment and maintenance.
Read how Deep Learning with R, Third Edition explains machine learning workflow →
The Art of AI Product Development
A machine learning lifecycle is an iterative process for preparing data, developing models, evaluating results, and applying machine-learning outputs.
- How it is used
- Guides teams from product questions through predictive modeling, evaluation, interpretation, and repeated improvement.
- How deep it goes
- Traces problem formulation, training data, algorithm selection, metrics, tuning, interpretation, monitoring, and iteration without detailing implementation internals.
- Examples
- Uses website recommendations, user segmentation, churn prediction, harmful-output checks, feedback ratings, and corrective training examples.
- What you can do after
- Evaluate product data through iterative predictive-AI work and refine models, monitoring, and activities in response to results.
Read how The Art of AI Product Development explains machine learning workflow →
LLMs in Production
A machine learning life cycle is the sequence of activities involved in creating, deploying, operating, and retiring machine-learning models.
- How it is used
- Frames machine-learning work as a life cycle rather than a collection of infrastructure tools.
- How deep it goes
- Contrasts life-cycle purpose with infrastructure mechanics but stops short of detailing each life-cycle phase.
- Examples
- Uses laptop-based scikit-learn systems, Jupyter Notebooks, and an LLM-oriented infrastructure diagram as concrete contrasts.
- What you can do after
- Evaluate machine-learning platforms by separating life-cycle needs from the infrastructure used to satisfy them.
Read how LLMs in Production explains machine learning workflow →
Related concepts
- pipelineequivalent to · A pipeline is an ordered or coordinated sequence of connected processing stages or tasks that passes data, work, or results toward an outcome.
- monitoringhas part · Monitoring is the ongoing observation and measurement of a system’s behavior, state, or operational signals to assess its condition and detect changes or problems.
- traininghas part · Training is a process of developing or improving a person’s or model’s capabilities through instruction, experience, or adjustment using data.
- preprocessinghas part · Preprocessing is the preparation or transformation of data, content, or other inputs into a form suitable for a subsequent computational operation.
- data ingestionhas part · Data ingestion is the process of acquiring or receiving data from one or more sources and bringing it into a system for storage or processing.
- model developmenthas part · Model development is the process of designing, constructing, training, evaluating, and refining a computational or statistical model for a defined task.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.