K-nearest neighbours (k-NN) is a simple supervised learning method. To classify a new data point, it looks for the k most similar known examples and adopts their most common class or, for numerical values, their average. It needs no real training phase, but becomes slow with large data sets and is sensitive to how features are scaled. Example: product recommendations based on similar customers.
A classifier is a model or method that assigns inputs to one of several predefined categories. Classic variants such as decision trees or naive Bayes are trained on labelled examples; language models can also sort texts based on a description of the categories. Typical uses include spam detection, sorting customer enquiries by topic and categorising receipts. Quality is measured with metrics such as accuracy, precision and recall.
K-means is an unsupervised learning method that divides data points into a predefined number k of groups. It assigns each point to the nearest group centre and keeps moving these centres until the assignments barely change. The method is fast and easy to understand, but requires a sensible choice of k and features on comparable scales. Example: grouping customers by purchase frequency and basket value.
In supervised learning, a model learns from examples for which the correct answer is already known, such as emails labelled “spam” or “not spam”. From these it derives rules to classify new, unseen cases or to predict values. Typical tasks are classification and regression; most of the effort goes into creating cleanly labelled data.
An embedding vector (embedding) is a list of numbers that represents the meaning of a word, sentence or document so that texts with similar content lie close together. This makes it possible to recognise, for example, that 'cancellation' and 'ending the contract' are related even though they share no words. Embeddings underpin many semantic search systems, recommendation services and RAG applications; they are computed by specially trained models.
