Starting and learning
Machine learning for non technical people
Machine learning learns a rule from labelled examples and applies it to new cases: scoring which customers are likely to respond, flagging faulty parts in camera images, or finding the passage in a document that answers a question. You can understand and try each of these without writing code.
How machine learning is used, explained without code
Learning from examples instead of rules
Instead of following instructions someone wrote, a model learns a rule from labelled examples. You can see this yourself by giving a few examples of each category and watching it sort new ones.
Real exampleTeachable MachineGoogle's Teachable Machine lets anyone train a computer to recognise their own images, sounds or poses, with no expertise or coding required.
Sorting and routing requests
Models read incoming emails or support tickets, predict the topic and send them to the right team or suggest a reply. Staff still make the final decision, but spend less time sorting.
Real exampleCOTA at UberUber researchers describe COTA, a system that classifies customer support tickets and selects suggested answers for support representatives using machine learning and natural language processing.
Turning speech into text
Speech recognition models convert audio into written words for meeting notes, captions and voice assistants. They learn from large amounts of audio paired with transcripts.
Real exampleWhisper speech recognitionOpenAI's Whisper paper describes speech recognition models trained on 680,000 hours of multilingual audio paired with transcripts collected from the internet.
Forecasting from past patterns
Forecasting models learn how past conditions led to outcomes, such as rainfall leading to river levels, and predict what comes next. Their value depends on acting on the forecast in time.
Real exampleGlobal prediction of extreme floods (Nature, 2024)Google researchers published an AI model that forecasts extreme river floods, including in watersheds that have no streamflow gauges.
Can non-technical people learn machine learning?
Yes. You can understand what machine learning does, judge whether a project is sound and train simple models with point-and-click tools without writing code. If you later want to build models yourself, Python is the next step and is learnable without a technical background.
- Core ideas in plain languageExamples, labels, training and test data, and why a model is judged on data it has not seen.Learn it in Machine Learning
- Reading results and errorAverages, percentages and error rates, and why a single accuracy figure can mislead when one outcome is rare.Learn it in Statistics
- Try a no-code modelTrain a small image or sound classifier in a browser tool, then test it on examples it has not seen.
- Optional: PythonIf you want to build models yourself, Python fundamentals let you follow and edit real notebooks.Learn it in Python
Moving from a non-technical role into machine learning
If you already work in the field
People in operations, marketing, policy or management already know the decisions and data their organisation has. Add the vocabulary of training, evaluation and error, then join a machine learning project as the person who defines the problem and checks the results.
If you come from data science or ML
Technical people working with non-technical colleagues need to learn how the decision is made today and what a wrong prediction costs. Spend time with the people who will use the model before building it.
Roles that do this workAI product managerBusiness analystAI programme manager
Tools you will use
| Tool | What you use it for |
|---|---|
| Orange Data Mining | Builds a full train-and-test workflow by connecting visual blocks, with no code required. |
| Teachable Machine | Trains an image or sound classifier in the browser from your own examples so you can see labels and mistakes. |
| KNIME Analytics Platform | Creates drag-and-drop data workflows for cleaning tables, training models and reviewing results. |
| Google Colab | Runs shared Python notebooks in the browser when you want to read or rerun someone else's analysis. |
Public data to practise on
- Iris: Measurements of three iris species, a small, clean table for seeing how a classifier separates categories.
- Palmer Penguins: Body measurements of three penguin species, a friendly table for classification and charts in no-code tools.
Your first project
Open the Iris dataset in Orange, train a decision tree to predict the flower species, and compare it with always guessing the most common species using accuracy on held-out flowers.
Common questions
Can I understand machine learning without maths?
Yes, at the level of what it does and how to judge it. You need examples, error and the difference between training and test data; the formulas matter when you start building models yourself.
What is the difference between AI and machine learning?
AI is the broad goal of making computers do tasks that seem to need intelligence. Machine learning is the main way that is done today: the system learns patterns from data rather than following rules written by hand.
Can I use machine learning without coding?
Yes. Browser tools such as Teachable Machine and no-code features in cloud and spreadsheet products let you train and test simple models. Coding gives more control and is needed for most production work.
Is ChatGPT machine learning?
Yes. It is built on a large language model, a neural network trained on large amounts of text to predict the next token and then further trained to follow instructions.
Courses for these skills
- Practical Machine Learning in PythonBuild the foundations for classification, anomaly detection and model evaluation.51 published lessons · more lessons in production
- Statistics and Probability for Machine LearningUnderstand base rates, sampling and uncertainty.41 published lessons · all planned lessons open
- Deep Learning with PyTorchStudy CNNs, transfer learning, object detection and segmentation in the Computer Vision syllabus.Syllabus available · lessons open as produced
- Generative AI: LLMs, RAG and AgentsStudy retrieval-augmented generation and evaluating grounded outputs.Syllabus available · lessons open as produced
- Data Tools: NumPy, pandas, Matplotlib and scikit-learnPreserve document identifiers, revisions and access metadata.23 published lessons · all planned lessons open
New to programming or maths? Start with Python, or follow the machine learning path in order. Published lessons are free to read; compare Free and Pro.