36 of the Computer Science glossary’s terms, each with the part of the syllabus where you first meet it.
The known correct output attached to an example in a dataset, which a supervised learning model learns to predict from the features.
A very large deep learning model, pre-trained on huge amounts of text, that processes and generates human language by repeatedly predicting the next part of a text.
The rule that the item added most recently is the first to be removed.
The delay between an event or request and the system's response to it.
A node with no children.
A graph of a model's error (or accuracy) on training data and on validation data plotted against the amount of training, the amount of data or the model's complexity, used to detect underfitting and overfitting.
A hyperparameter that sets the size of the change made to a model's weights at each step of training.
The method of fitting a regression line by choosing the slope and intercept that minimise the sum of the squared residuals.
A collection of pre-written, tested modules that programs can import to use their subprograms instead of writing the code themselves.
For a rule X → Y, the confidence of the rule divided by the support of Y; a lift above 1 means X and Y occur together more often than if they were independent, 1 means no association, and below 1 means a negative association.
An open addressing method that, when the calculated index is occupied, tries index + 1, index + 2 and so on (wrapping round to 0 at the end of the table) until it finds a free bucket.
A supervised learning method that models the relationship between one or more input variables and a continuous output variable with a straight line (or flat surface), and uses that line to predict the output for new inputs.
A search algorithm that checks each element of a list in order, from the first, until it finds the target or reaches the end of the list.
A complexity in which the number of steps grows in direct proportion to the input size, so that doubling the input doubles the work.
A complexity in which the number of steps grows in proportion to multiplied by .
Describes two classes of data that can be completely separated by a straight line (or, in more dimensions, a flat plane).
A table that resolves a many-to-many relationship between two tables by holding their primary keys as foreign keys, usually combined to form its own composite primary key.
A dynamic data structure that stores a sequence of items in nodes, where each node holds a data item and a reference to the next node in the sequence.
Python's built-in ordered, mutable and dynamic sequence, indexed from 0, which grows and shrinks as elements are added and removed.
Spreading incoming requests across several servers that provide the same service, so that no single server is overloaded.
The number of entries stored in a hash table divided by the number of buckets, .
A network that connects devices within a small geographic area, typically one building or site, and is usually owned and managed by a single organisation.
A variable declared inside a subprogram or block, which can only be used there and exists only while that code is running.
The tendency of programs to access the same memory locations repeatedly within a short time (temporal locality) and to access locations close to ones recently used (spatial locality).
The property of a distributed database that lets users query data without knowing or stating at which site or node it is stored.
A marker that a transaction places on a piece of data (a record, a page or a whole table) so that other transactions cannot change it, or sometimes even read it, until the lock is released.
The power to which a base must be raised to give a number, so that means ; in computer science,
A complexity in which the number of steps grows in proportion to the logarithm of the input size, so that doubling the input adds only a constant amount of work.
A diagram of a logic circuit that uses standard gate symbols connected by lines (wires) to show how the inputs are processed to produce the outputs.
An error in the design or code of a program that makes it produce incorrect results or behaviour without stopping it or producing an error message.
An electronic circuit that takes one or more binary inputs and produces a single binary output determined by a Boolean operation.
A description of the data as relational tables, giving every field with a generic data type, the primary and foreign keys and the integrity constraints, without detail of how a particular DBMS stores the data.
A function that measures how far a model's predictions are from the correct values, giving a single number that training tries to make as small as possible.
Compression in which no information is lost, so decompressing the data restores an exact copy of the original.
Compression that permanently removes information that people are unlikely to notice, so the decompressed data is only an approximation of the original.
A programming language that corresponds closely to the instruction set of one processor family, so that each statement usually stands for a single machine instruction.