Concepts starting with S

381 concepts, each with a plain-language explanation, examples and diagrams. Pick a letter or open a concept.

Sample Complexity — Regression loss functions measure more than whether a prediction is wrong: they quantify the discrepancy between prediction and…
Sample Complexity of Finite Hypothesis Classes — Three conjunctions connected by OR define a 3-term DNF.
Sample Complexity of Neural Networks — VC dimension is a measure of hypothesis-class capacity based on shattering.
Sample Model — Planning uses model-generated experience, while learning uses experience from the real environment.
Sample Models in Blackjack — The key distinction is whether the model exposes the full probability description or only one sampled outcome.
Sample Size Flexibility — The two notions share the same competitiveness requirement.
Sample-Average Action-Value Estimation — A sample average is a natural estimate of an action value.
Sample-Average Action-Value Methods — Initial estimates affect action-value methods because they provide the starting values used before experience is available.
Sample-Average Method and Nonstationarity — Sample averages are challenged when the values they estimate do not remain fixed.
Sampling from Probability Distributions — Character-level generation learns the relationship between a fixed-length character context and the character that follows it.
Sampling Production Traces — The per-trace loop is the daily cycle the FDE works inside: production traces are sampled, a pattern is identified, a fix is…
Sampling Random Tokens with torch.multinomial — The modified generate function adds top-k filtering and temperature scaling to the earlier loop.
Sarsa — Expected Sarsa is a Q-learning variation based on an expected next-state value.
Sarsa algorithm — Calculate the n-Step TD Error before applying the semi-gradient Sarsa update.
Sarsa and n-step Sarsa — Differential value functions are the average-reward counterparts of familiar value functions.
Sarsa and On-Policy TD Control — TD control joins value prediction with local policy improvement.
Sauer's Lemma — The proof has separate lower-bound and upper-bound routes.
Saving a GPT Model and Optimizer Checkpoint — Loading a GPT model and optimizer from a checkpoint is a two-part act: the program constructs live objects first, then attaches…
Saving a PyTorch Model — torch.save(model.state_dict(), file_name) writes a model's state dictionary to a .pth file, so the fine-tuned model can be…
Saving and Loading Keras Models — A Keras callback is an object passed to fit that is called at various points during training.
Saving and Loading LLM Weights — A weight parameter is a model parameter that is learned during training, not typed in by hand.
Saving and Loading Model Weights — Training loss measures fit on the training data, and validation loss measures fit on held-out data; both should be read together.
Saving and Loading Model Weights in PyTorch — Temperature and top-k are settings that control the randomness of text generation, and they do so in different ways.
Saving and Loading PyTorch Model Weights — A saved model can be resumed: load the model and optimizer in a new Python session or notebook, then continue pretraining for…
Saving and Loading PyTorch Models — torch.save() writes data to a file. In the source example, the data is model.state_dict() and the file is "model.pth".
Saving Model Weights in PyTorch — The recommended way to save a PyTorch model is to save its state_dict, passed as the first argument to torch.save.
Saving PyTorch Models — torch.load("model.pth") reads the saved file and reconstructs the Python dictionary containing the model's parameters.
Scaled Dot-Product Attention with Trainable Weights — Self-attention creates three views of each input token by multiplying it with three trainable weight matrices: W_q for queries…
Scaling Images — The Graphics class provides the methods that copy an image onto the screen or an off-screen canvas, most commonly from inside…
Scope and Namespace in Python — Built-in function names like max, len, sum, and print should be treated as reserved words and never used as variable names.
Scope and Variable Lifetime in Functions — When you declare variables inside a function definition, they are not related in any way to other variables with the same names…
Scope and Variable Visibility — Functions allow you to name groups of statements, which is the foundation of readable code.
Scope Management in Forward Deployed Engagements — Practicing conversational responses means rehearsing specific words for situations that can be anticipated, so they are…
Scope: Local and Global Variables — Expressions and variables can both be passed as arguments to user-defined functions, just as they can be passed to built-in…
Scoping an Engagement with the Customer — The customer outcome metric is the measure agreed in scoping and the one you are graded on internally.
Scoping and Validation Before Delivery — The delivery phase deploys a developed system to a live environment, and it goes well when scoping and validation have been done…
Scoping Data Work in AI Deployments — A clean sample is evidence about the sample; production data can differ, and the source reports exactly that gap.
Scoping Discipline and Documentation — A scoping document is a one-page record of the shared understanding between customer and vendor, and its format matters less…
Scoping Discipline in Deployment Work — Demand for deployment FDEs is in the tens of thousands of roles, and current supply is only a fraction of that, so deployment…
Scoping in AI Projects — The Validation Phase is the second phase of the FDE motion and validates the scoped thing before the team commits to delivery.
Scoping in Consulting Engagements — Standard consulting moves from scoping directly to a statement of work and a delivery timeline.
Scoring Model Responses — generate_model_scores builds a prompt for each entry, asks a model for a 0 to 100 score, converts the reply to an integer, and…
Search in Game-Playing Programs — A value function may be internally consistent without being useful for choosing moves.
Searching a List with Linear Search — The List interface extends Collection and adds methods that act on items by numerical position.
Searching an Array — Whole-array problems are solved by visiting every element with a for loop and combining what you find.
Searching an Array for a Value — An index equal to an array's length is illegal, because the legal indices run from 0 to length - 1.
Searching and Filtering Strings — Traversal is the process of handling a string one character at a time, from beginning to end.
Searching and Pattern Matching in Strings — A sequence is an ordered set of values identified by integer indices starting at 0; strings are sequences of characters.
Searching the Space of Policies — Evolutionary methods evaluate a fixed policy through many game outcomes.
Second Normal Form: Removing Partial Dependencies — Redundancy occurs when the same information is stored multiple times across different rows in a table, such as repeating an…
Secondary Reinforcement — Eligibility traces help reinforcement learning algorithms address delayed reinforcement.
Secondary Reward — A reward signal is a numerical evaluation, not necessarily a physical reward.
Secondary Storage: Permanent Data — Input devices capture user actions and send data into the computer; output devices receive data from the computer and present it…
Secure Credential Storage and Environment Variables — API keys are the mechanism vendors use to identify users and monitor their service consumption.
Securing API Access with OAuth — OAuth libraries are free, pre-built tools that handle the complexity of the OAuth 2.0 protocol so you do not have to implement…
SELECT Statements and Query Basics — JOIN reconstructs data from multiple normalized tables by matching rows based on key values specified in the ON condition.
SELECT Statements: Retrieving Data from Tables — SQL uses a single equal sign (=) for equality testing in WHERE clauses, not the double equal sign (==) used in Python.
SELECT: Retrieving Data from Tables — UPDATE modifies existing records in a table using the SET keyword to specify columns and new values.
Selecting Trainable Layers — Fine-tuning adjusts a pre-trained model's weights to fit a specific task, and it can either update the whole model or only…
Selection Sort — A shuffling algorithm aims to put the elements of an array into a random order, with the typical case being shuffling a deck of…
Self-Attention in Transformer Language Models — A causal attention mask is used when an LLM reads and generates text from left to right, to prevent access to future tokens.
Self-Attention in Transformers — The Bahdanau attention mechanism lets the decoder selectively access different parts of the input sequence at each decoding step.
Self-attention with trainable weights — All context vectors are computed with one matrix multiplication: attn_weights @ inputs.
Self-Play — TD-Gammon combined self-play with TD learning to improve its backgammon performance.
Self-Play Learning Methods — TD-Gammon's importance came from learning strong play with little backgammon knowledge.
Semi-gradient methods — Episodic Semi-gradient Sarsa learns a value function by updating value-function weights θ during episodes.
Semi-gradient Methods for Value Prediction — SGD performs an immediate, small weight-vector adjustment for each example.
Semi-gradient n-Step Sarsa — Differential semi-gradient n-step Sarsa extends semi-gradient n-step Sarsa with differential reward estimation.
Semi-gradient prediction methods — Action-value prediction changes the training input from a state to a state-action pair.
Semi-Gradient Q-Learning — Semi-Gradient Expected Sarsa is a one-step algorithm for action values.
Semi-gradient Sarsa — Calculate the n-Step TD Error before applying the semi-gradient Sarsa update.
Semi-Gradient TD(0) — SGD variations are used to find a good weight vector for function approximation.
Sending XML Data Over HTTP — Triple single quotes (''') and triple double quotes (""") both create multi-line strings in Python.
Sentiment Analysis — Sequence data must be represented numerically before a deep-learning model can process it.
Sentinel Values in Input Processing — A while loop can read and process integers repeatedly until a condition becomes false.
Sentinel-Controlled Input — Priming a loop means preparing a valid current value for the first while-condition test.
Sequence Data — Timeseries data consists of measurements arranged in sequence at regular time intervals.
Sequence Modeling — A language model learns the statistical structure of language by modeling possible next tokens.
Sequence Processing Applications — Sequence models operate on ordered numeric representations, not raw text.
Sequence Processing with Convnets — Deep learning applications can be organized by their input and output modalities.
Sequence Processing with Recurrent Layers — Multi-input models use separate branches for separate data sources.
Sequences of Training Examples — Lowercase letters commonly denote scalars or abstract objects; boldface lowercase letters emphasize vectors.
Sequential Decision-Making — Bandit problems have a history spanning statistics, engineering, and psychology.
Sequential Models in Keras — The Functional API describes a model as a connected route of tensor transformations.
Sequential neural-network models — The example creates a Keras Sequential model and adds layers in a fixed order.
Sequential Processing of Arrays — Sequential access processes array elements one after another in array order, while random access lets you reach any element…
Sequential Streams — A collection serves as the source of values for a stream, and this applies to lists, sets, and any other collection.
Setting Component Backgrounds — BorderFactory.createEmptyBorder(top, left, bottom, right) creates an empty border that reserves space around the edges of a…
Setting Seeds for Reproducible Randomness — random.random() generates a pseudorandom float in the range [0.0, 1.0), inclusive of 0.0 but exclusive of 1.0
Setting Up a Deep-Learning Workstation — A recent NVIDIA GPU is recommended for faster execution, not required for running the examples.
Setting Up Java Programming Environments — This lesson explains how to begin the Java course rather than teaching a full programming concept.
Setting Up the Java Development Kit — jshell provides a direct command-line way to try Java code.
Setting Up Your Python Environment — For Windows 2000, XP, 2003 , click on Control Panel → System → Advanced → Environment Variables . Click on the variable named…
SGD and Backpropagation for Neural Networks — Backpropagation calculates the gradient of a loss with respect to network weights for an example.
SGD and subgradients of maximum functions — Multiclass SVM handles several labels by scoring input-label pairs and selecting the largest score.
Shallow Copy Versus Deep Copy — Aliasing occurs when two or more variables refer to the same object in memory. The assignment b = a creates an alias, not a copy.
Shallow vs. Deep Copying — Identity (is) checks whether two variables refer to the same object in memory; equivalence (==) checks whether two objects have…
Shapes and Outlines with drawRect and drawOval — Aliasing is the jagged appearance of shapes drawn on a pixel grid, and antialiasing is the technique that reduces that…
Shared Resources in Multithreaded Programs — Use synchronization when multiple threads share a resource and their access must be exclusive.
Shattering — Intervals over the real numbers form a class of binary-valued functions.
Short-Circuit Evaluation in Logical Operators — Python allows non-boolean values as operands in logical operators, interpreting any nonzero number as 'true'
Showing and Hiding Windows with setVisible — The default close operation, set with setDefaultCloseOperation, decides what happens when the user clicks a JFrame's close box.
Shuffling Lists with random.shuffle() — random.randint(low, high) generates a random integer within an inclusive range, including both the low and high endpoints.
Side Effects vs. Return Values — Void functions perform actions but return None, not a usable value.
Sigmoid Function — Logistic regression is used for classification tasks.
Similarity Measures — Spectral clustering begins by representing relationships between data points as a weighted similarity graph.
Simple Generic Classes — A type parameter is a placeholder written in angle brackets after a class name, and it works much like a formal parameter in a…
Simple if Statements — A multiway branch built from if, else if, and else executes one and only one of its statements.
Simple Pricing Models — A product's deployment cost model describes the costs of setting up and integrating it, and some products have no deployment…
Simple Self-Attention Without Trainable Weights — SelfAttention_v2 builds its query, key, and value projections with nn.Linear layers, configured by d_in, d_out, and qkv_bias.
SimpleRNN networks — The Keras example uses an LSTM layer in a model trained on IMDB data.
Simplified Self-Attention with Dot Products — Weight parameters are learned coefficients optimized during training, while attention weights are dynamic, context-specific…
Simplified Self-Attention Without Trainable Weights — Self-attention computes attention weights by relating different positions within a single input sequence, while traditional…
Simplifying Conditionals with Logical Operators — A nested conditional is a conditional statement placed inside another conditional statement, typically within one of the…
Simulated Transitions and Backups — Uniform planning is easy to describe but can waste many backups.
Single Linkage — Clustering is an unsupervised organization problem rather than a labeled prediction problem.
Single Linkage Clustering — The supplied source pack does not define PCA or variance maximization.
Single-Head Attention and Stacked Attention Layers — MultiHeadAttention uses one projection per role (query, key, value) to a width of d_out, then reshapes that width into num_heads…
Single-Head Attention Modules — Multi-head attention extends causal attention across several heads, each operating independently with its own weights.
Singular Value Decomposition — A symmetric matrix is positive definite when every eigenvalue is positive.
Skipping Iterations with continue — A while loop is a looping statement that repeatedly executes code as long as a condition is true.
Slack Variables — Soft-SVM combines labeled training pairs and a positive λ parameter in an optimization problem.
Slicing Lists — pop removes an element by index and returns its value; use it when you need to capture the removed element.
Slicing Lists to Extract Subsets — Indices start at 0 and can be positive (counting from the start) or negative (counting from the end).
Slicing Strings — A string is a sequence of individual characters that can be accessed one at a time using the bracket operator.
Slicing with Steps and Negative Steps — A slice with start >= end always produces an empty string, not an error.
Sliding Window Approach for Generating Input-Target Pairs — An LLM is pretrained by predicting the next word in a text, so training data is built as input-target pairs.
Socket Programming Basics — Binary files must be accumulated in a bytes buffer and cannot be printed as they arrive like text files.
Socket Programming Fundamentals — When you call encode() on a string, Python looks up each character and converts it to one or more bytes according to the UTF-8…
Sockets and Client-Server Communication — A threaded server accepts each connection in its main loop and hands that connection to a new thread rather than handling it…
Sockets and Network Connections — A socket is one end of a network connection and has an input stream for incoming data and an output stream for outgoing data.
Soft-max Action Selection — Numerical preferences express how strongly each action is favored.
Soft-SVM and Regularization — Soft-SVM combines labeled training pairs and a positive λ parameter in an optimization problem.
Softmax and Converting Logits to Probabilities — Probabilistic sampling picks the next token with a chance proportional to its probability, so a most-likely token is chosen most…
Softmax and Probabilities for Token Selection — The modified generate function adds top-k filtering and temperature scaling to the earlier loop.
Softmax and Probability Scores — Argmax converts a row of model output into a class label by returning the position of the largest value.
Softmax Function and Attention Weights — A causal attention mask zeroes out attention weights for future positions so that each position attends only to itself and…
Sorted Arrays — Binary search finds an item in a sorted array by inspecting the middle of a shrinking range, and each comparison eliminates…
Sorting Algorithms Overview — Unsorting is the reverse goal of sorting: it aims for a random arrangement rather than an ordered one.
Sorting and Ordering Data — Use a dictionary to count word frequencies by checking if each word exists and incrementing its count
Sorting and Ordering Data in Python — max() and min() are built-in functions that find the largest and smallest values in any collection.
Sorting and Ordering Data Structures — Finding unique words involves a pipeline: read file → split lines into words → check membership to avoid duplicates → sort…
Sorting and the key Parameter — A lambda statement creates an anonymous function object with one parameter and one expression.
Sorting and the sort() Method — Lambda functions used as sort keys provide a temporary comparison value for each element without modifying the original list.
Sorting Dictionaries by Value — urllib.request.urlopen() opens a web URL and returns a file-like object that you can iterate over line by line.
Sorting Dictionary Data — Use a dictionary to count word frequencies by checking if each word exists as a key and either initializing it to 1 or…
Sorting Dictionary Keys — Looping through a dictionary directly iterates over its keys, not its values.
Sorting Dictionary Output by Frequency — Nested loops are essential for processing multi-level data structures like lines and words in a file. The outer loop iterates…
Sorting Dictionary Results — Dictionaries outperform lists and variables for character counting because they only allocate space for characters that actually…
Sorting Lists and Other Sequences — The sort function compares tuples element-by-element, starting with the first position and moving to subsequent positions only…
Sorting Lists in Python — To print dictionary keys in alphabetical order, extract keys using list(dict.keys()), sort the resulting list with sort(), and…
Sorting Lists of Tuples — Python compares tuples element-by-element from left to right, stopping at the first index where elements differ.
Sorting Lists with .sort() and sorted() — Dictionaries do not guarantee a predictable order, so you convert them to a list of tuples using list(d.items()) to enable…
Sorting Lists with the sort() Method — Tuple comparison proceeds element by element from left to right, stopping as soon as a difference is found.
Sparse Models — The ℓ1 constraint restricts the hypothesis class through ‖w‖1 ≤ B and is associated with sparse models.
Sparse Recovery — RIP is a property of a matrix W, not a property of an individual vector.
Sparse Representations — Dictionary learning creates a feature vocabulary for data that may not have an obvious vocabulary.
Sparse Vectors — RIP is a property of a matrix W, not a property of an individual vector.
Sparse Vectors in Compressed Sensing — RIP is defined through a restriction involving nonzero vectors with ‖x‖0 ≤ s.
Spatial Hierarchies in Computer Vision — Max pooling replaces each local window with its maximum value.
Special Cases in Reinforcement Learning: Afterstates and Games — An afterstate describes the task immediately after an action.
Special Context Tokens — Tokenization cuts text into individual tokens, and a new text sample is cut in the same way before any numbers are assigned.
Special Methods and Operator Overloading — The class keyword creates a new type in Python, not just a container for code.
Spike-Timing-Dependent Plasticity — STDP makes synaptic change sensitive to the order and timing of neural spikes.
Splitting Text into Word and Subword Tokens — Large language model training requires text to be prepared first, and the source treats this as a necessary step.
Splitting the Dataset into Training, Validation, and Test Sets — The to_csv method, called on a pandas DataFrame, saves that DataFrame to a CSV file named by its first argument.
SQL Fundamentals: SELECT, INSERT, UPDATE, DELETE — The Database Browser for SQLite is a visual tool that reads SQLite database files and displays their contents in a user-friendly…
SQL INSERT, SELECT, and WHERE Clauses — Many-to-many data loading requires parsing JSON and distributing the data across two entity tables and one junction table.
SQL INSERT, SELECT, and WHERE Statements — The INSERT OR IGNORE and SELECT pattern solves the foreign key dependency problem by inserting a record and immediately…
SQL JOIN Clauses and Data Reconstruction — The INSERT OR IGNORE and SELECT pattern is a two-step sequence that inserts a row and immediately retrieves its primary key for…
SQL Queries and Data Retrieval — A database is justified only when your application needs frequent random updates on large datasets, must handle data larger than…
SQL SELECT, INSERT, and UPDATE Statements — The INSERT OR IGNORE and SELECT pattern is a two-step sequence that inserts a row and immediately retrieves its primary key for…
Squared Loss — For linear regression with squared loss, least squares solves the empirical risk minimization problem.
Stability Issues with Other Methods — Averagers prevent instability by avoiding extrapolation from observed targets.
Stable Learning Rules — Tikhonov regularization augments the loss with λ ‖w‖2.
Stack Machine Code for Expressions — A BinOpNode stores an operator and two child expression nodes, and its value is computed by applying the operator to the values…
Stack Machines and Machine Language — An expression tree is a data structure that represents an expression, with operators as internal nodes and numbers as leaves.
Stack Operations — A stack is a sequence of items piled one on top of the other, and only the top item is accessible at any time.
Stack Operations and the LIFO Principle — A stack can be implemented with an array or with a linked list, and both versions expose the same push, pop, and isEmpty…
Stacking Multiple Attention Layers — Multi-head attention extends causal attention across several heads, each operating independently with its own weights.
Stacking Single-Head Attention Layers — The current MultiHeadAttentionWrapper processes its heads sequentially through a list comprehension in the forward method.
Stacks and Their Operations — A postfix expression places each operator after the two operands it acts on, so the order of symbols alone determines the…
Stacks: Push and Pop Operations — A stack-based evaluation algorithm can detect errors in a postfix expression without needing parentheses or a human reader.
Staff-Track and Senior-Track Careers — A pattern becomes a product feature through four stages: recurring engagements, a selected pattern, a generalized version…
Stage 3: Fine-Tuning the Pretrained LLM — Stage 3 of the course takes a pretrained LLM and fine-tunes it for specific work.
Stakeholder Alignment in AI Projects — A proposed AI project should move past scoping only if the problem is among the customer's top five priorities.
Stakeholder Management in Regulated Industries — In regulated industries, trust is the gating factor that controls the data customers will share, the workflows they will allow…
Stakeholder Maps and Discovery Techniques — The source states that the single most common reason an FDE engagement fails is not technical, but picking the wrong problem to…
Stakeholder Roles in Manufacturing Projects — Stakeholders include anyone who can affect or be affected by a project, which usually goes beyond the technical team.
Standard Array Methods — Arrays.copyOf(list, lengthOfCopy) returns a new array of the requested length, copying items from the original and filling any…
Standard Dialog Boxes in Java — A custom dialog box is built as a subclass of JDialog, and it is a separate window much like a frame.
Standard Input and Output in Java — The TextIO class is a class in the textio package that contains static methods for input and output operations.
Standard Input with TextIO — TextIO can redirect input from standard input to a file.
Standard Reinforcement Learning Concepts — The section broadens reinforcement learning beyond its standard ideas by examining relationships with psychology and neuroscience.
State Aggregation — The history of function approximation in reinforcement learning includes gradient-descent methods, bootstrapping, state…
State and Action Representations — Afterstate value functions assess a state after an action has been taken.
State and Action Representations in Reinforcement Learning — An afterstate describes the task immediately after an action.
State Representation — Tile coding converts continuous states into coarse, tile-based features.
State Representation in Reinforcement Learning — A state signal may combine immediate sensations with information derived from past sensations.
State Spaces — The number of states often grows exponentially as the number of state variables increases.
State Transitions and Environment Models — A cognitive map captures learned relationships among successive stimuli.
State Value Functions and Action Values — Watson treated Daily-Double wagering as a decision under uncertainty rather than as a fixed betting rule.
State Values — An action value evaluates a state together with a particular action.
State Values and Action Values — A transition graph helps organize how an MDP moves from a state and action to possible next states and rewards.
State Visitation in Reinforcement Learning — An episodic on-policy distribution depends on both episode starting states and transitions within episodes.
State-Action Associations — Cognitive maps and environment models are representations of environments, described respectively in psychological and…
State-Action Value Functions — The Bellman equation is a recursive consistency relationship for value functions.
State-Space Planning — Incremental planning breaks planning into small steps instead of requiring one uninterrupted calculation.
State-Space Planning and Random-Sample Q-Planning — Random-sample Q-planning begins by selecting S and A at random.
State-Value and Action-Value Functions — An optimal value function evaluates the best achievable outcome from a state.
State-Value Backups — Asynchronous DP does not follow systematic sweeps of the state set.
State-Value Estimates — Bootstrapping is the practice of revising an estimate on the basis of another estimate.
State-Value Estimation under a Policy — The methods differ only in how they select returns from episodes with repeated visits.
State-Value Estimation with Monte Carlo Methods — The target of estimation is qπ(s, a): the expected return after taking a in s and then following policy π.
State-Value Functions — Optimality is evaluated across all states, not from one state alone.
State-Value Prediction — TD(0) and Monte Carlo can disagree because they use different evidence from the same episodes.
State-Visit Distributions — The policy gradient theorem links performance changes to the policy-weight vector θ.
Statements vs. Expressions — An expression is a combination of values, variables, and operators, or a single value or variable by itself.
States and Actions in Reinforcement Learning — An afterstate is the state reached immediately after an action.
States, Actions, and Rewards — Episodic interaction has natural boundaries and restarts its time numbering in each episode.
States, Actions, and Rewards in Reinforcement Learning — MDP notation distinguishes individual objects from sets of possible objects.
Static and Instance Members — A simple name inside a method is resolved first against local variables and parameters in scope, and only then against member…
Static and Non-Static Members in Classes — A class is a blueprint, or factory, used to construct objects; it describes what objects will contain rather than being the…
Static and Non-Static Subroutines — Drawing instructions belong inside a subroutine definition, and that subroutine belongs inside a class definition.
Static Member Variables — A static nested class is a class declared inside another class with the static modifier, and it is part of the static structure…
Static Member Variables in Classes — A subroutine can use local variables declared inside it, formal parameters that receive values from the caller, and variables…
Static Members and Class Variables — A named constant gives a meaningful name to a value that is meant to stay fixed for the whole program.
Static Members and Shared Class Data — A class definition is a blueprint that declares the instance variables and instance methods every object of that class will…
Static Members of a Class — A static nested class has full access to the static members of its containing class, including private ones, and the containing…
Static Methods and Class-Level Members — A static nested class is a class declared inside another class with the static modifier, and it is part of the static structure…
Static Methods and the Class Name in Java — A GUI program opens a window on the screen and interacts with the user through it, and a simple one can be written in only a few…
Static Methods and Their Definitions — A static method is called from another class by writing the class name, a dot, and the method name: ClassName.methodName();
Static Methods in Java — The standard array-processing methods are static methods defined in the Arrays class, which is in the java.util package.
Static Nested Classes — Compiling a class with a nested class creates two separate class files.
Static Nested Classes in Java — A non-static nested class is called an inner class, and it is associated with an object of the containing class rather than with…
Static Subroutines and Variables in Java — Modifiers appear at the front of a subroutine definition and set characteristics such as whether it is static or public.
Static Subroutines in Java — A member variable is declared outside any subroutine but inside a class; a static member variable belongs to the class as a…
Static Variables and Methods — Functions return values that calling code can use.
Static Variables and Static Initialization — An instance variable can receive an initial value directly in its declaration, such as public int die1 = 3;, and that value is…
Static Variables and Static Methods — A static import lets a static member of a class be referred to by its simple name, which removes the need for compound names…
Stationary and Nonstationary Bandit Problems — High initial action-value estimates can make a greedy method explore without explicitly using ε-greedy random selection.
Stationary Distributions and Markov Transition Matrices — Linear TD(0) relies on the matrix A to control whether weight updates shrink or grow.
Statistical Error Bounds — The learning protocol separates hypothesis construction from performance estimation.
Statistical Learning Framework — X names the collection of objects being studied, while Y names the collection of possible labels.
Statistical Learning Framework Fundamentals — Getting Started introduces how to approach the course rather than requiring you to memorize its entire structure.
Statistical Learning Theory — Bandit problems have a history spanning statistics, engineering, and psychology.
Statistics and Optimization in Reinforcement Learning — Reinforcement learning is connected to both mathematical disciplines and the sciences of biological learning.
Staying Current with LLM Research — arXiv is a source for recent research papers on LLMs, including the recent listing for cs.LG.
Steady-State Distribution — Continuing problems have no episode endpoints, so policy performance is measured by average reward per time step over the long…
Step-Size Parameters — A suitable step-size sequence must satisfy two requirements: it must be large enough initially and small enough eventually.
Step-Size Parameters in Reinforcement Learning — Sample-average updating reduces the step size according to αn(a) = 1/n.
Stimulus Control and Behavior Modification — Terminology must be interpreted within the field that uses it.
Stimulus Representation — The TD model places time and changes in prediction at the center of its account of classical conditioning.
Stimulus Traces and Secondary Reinforcement — A stimulus trace persists in the nervous system after a stimulus ends.
Stimulus Traces in Instrumental Conditioning — A delayed primary reward does not have to be the only influence operating during the delay.
Stochastic and Semi-gradient Methods for Value Prediction — SGD applies a sequence of parameter updates to function approximation in value prediction.
Stochastic Approximation in Reinforcement Learning — The gradient bandit algorithm can be understood as a stochastic approximation to gradient ascent.
Stochastic Environments — Incorrect models give planning an inaccurate description of the environment.
Stochastic Gradient Descent — The gradient identifies the direction in which the function increases fastest around the current vector.
Stochastic Optimal Control — The historical problem is sequential decision-making under uncertainty, not isolated one-time choice.
Stochastic Policies — Softmax action selection over action preferences can approach deterministic behavior when the optimal action's preference is…
Storage Devices — Binary patterns encode machine language instructions and data.
Storage Devices and File Systems — Main memory is fast and nearly as quick as the CPU, but it is volatile and loses all data when power is cut off.
Storing and Displaying User Data with Variables — Use input('prompt string') to display a message before pausing for user input
Storing and Querying Spider Data with SQLite — spdump.py reads spider.sqlite and displays pages in a five-element tuple format: (incoming_links, old_rank, new_rank, page_id…
Storing Data in Memory — A variable is a named memory location that can hold data.
Storing Data in SQLite Databases — Responsible spidering means retrieving data at a controlled rate to respect server resources; gmane.py retrieves one message per…
Storing Drawings as Lists of Shape Objects — An off-screen canvas is a BufferedImage, a subclass of Image, that stores a picture in memory instead of showing it directly.
Storing Geographic Data in Databases — SQLite serves as a local cache to store geocoding results, preventing redundant API calls to the same locations.
Storing Values in Variables — Go to the end of the line under Variable value and append ;C:\Python27 (please verify that this folder exists, it will be…
Striatum — The basal ganglia are a connected collection of nuclei that can influence movement, decisions, learning, and reward processing.
String Basics — An example of a literal constant is a number like 5 , 1.23 , or a string like 'This is a string' or "It's a string!" .
String Basics and Indexing — Strings are objects belonging to the str class and possess methods for checking, searching, and transforming text.
String Comparison and Equality in Python — An Easter Egg is a harmless, hidden feature programmers add for fun; in this exercise, it is a special message triggered by the…
String Concatenation — The format() method constructs strings by substituting placeholders with argument values, replacing the need for error-prone…
String Concatenation and Combining Text — str() converts any value into its string representation without modifying the original value
String Concatenation and Conversion — input() pauses your program and waits for the user to type something on their keyboard, then returns what they typed as a string.
String Concatenation and Repetition — The input() function returns a string, so you must use int() to convert it to an integer for arithmetic operations.
String Concatenation with the + Operator — The * operator repeats a string by multiplying it by an integer, producing a new string with the original string concatenated…
String Conversion and Type Casting — but that is much uglier and error-prone. Second, the conversion to string would be done automatically by the format method…
String Data Type and Literals — The + operator performs concatenation when used with strings, joining them end to end to create a new string.
String Escape Sequences — Newline characters mark the end of lines in text files, but different systems use different representations: \n, \r, or \r\n.
String Escape Sequences in Python — Triple single quotes (''') and triple double quotes (""") both create multi-line strings in Python.
String Formatting — A string is a sequence of characters, ordered by position, and is one of the most frequently used data types in Python.
String Formatting and the format() Method — The format() method constructs strings by substituting placeholders with argument values, replacing the need for error-prone…
String Formatting in Python — Invisible characters like spaces, tabs, and newlines can cause silent errors in file operations.
String Immutability — String methods are built-in functions that transform, search, and analyze text. Each method behaves differently and returns a…
String Immutability and Method Chaining — translate() with str.maketrans() is the most flexible method for removing or replacing characters; use empty fromstr and tostr…
String Indexing and Access — A sequence is an ordered set of values identified by integer indices starting at 0; strings are sequences of characters.
String Indexing and Accessing Characters — The find() method locates the position of a substring and returns its starting index; use find(substring, start_pos) to search…
String Indexing and Negative Indices — A slice with start >= end always produces an empty string, not an error.
String Indexing Fundamentals — Use find() to locate a delimiter like a colon within a string, which returns the index of that character.
String Indexing with Integer Positions — A slice extracts a substring using [n:m], where n is the starting index (inclusive) and m is the stopping index (exclusive).
String Iteration and Indexing — The counter pattern is a fundamental computational technique: initialize a counter to 0, loop through data, check a condition…
String Literals and Constant Values — Text blocks provide a readable Java 17 syntax for multiline strings.
String Literals and Delimiters — An example of a literal constant is a number like 5 , 1.23 , or a string like 'This is a string' or "It's a string!" .
String Literals and Escape Sequences — An example of a literal constant is a number like 5 , 1.23 , or a string like 'This is a string' or "It's a string!" .
String Manipulation and Conversion — int() converts values to integers by truncating (not rounding) the decimal portion
String Manipulation in Python — A palindrome is a text that reads the same forwards and backwards, identified by comparing the original string to its reversed…
String Methods and Formatting — The + operator joins strings end-to-end (concatenation), not mathematical addition.
String Methods and Immutability — Most list methods modify the list in place and return None, unlike string methods which return a new string. This is because…
String Methods and Indexing — Email headers follow a predictable format; the day of the week is always the third word (index 2) on lines starting with 'From'.
String Methods and Manipulation — Strings are objects belonging to the str class and possess methods for checking, searching, and transforming text.
String Methods and Splitting — Use the continue statement to skip lines that do not meet your criteria, jumping immediately back to the next iteration of the…
String Methods and Text Processing — The file search pattern reads through a file line by line, applies a filter condition using string methods, and processes only…
String Methods and Transformations — A sequence is an ordered set of values identified by integer indices starting at 0; strings are sequences of characters.
String Methods for Searching and Replacing — A slice extracts a substring using [n:m], where n is the starting index (inclusive) and m is the stopping index (exclusive).
String Methods Overview — Square brackets in documentation indicate optional arguments; they are not part of the actual code.
String Methods: find() and in Operator — startswith() checks if a string begins with a specific prefix and returns True or False—it does not search for the substring…
String Methods: lower, strip, and replace — Method invocation uses dot notation: object.method_name() instead of function syntax like function_name(object).
String Methods: lower() and upper() — The Python split() function preserves punctuation attached to words, treating 'soft!' and 'soft' as different tokens and…
String Methods: replace() and split() — String methods like find(), strip(), startswith(), and lower() are essential tools for text processing in Python.
String Methods: split, find, startswith — A histogram is a frequency count stored in a dictionary, where keys are unique values (like email addresses) and values are…
String Methods: split() and join() — Unpacking is a syntactic feature that assigns multiple variables from a sequence in one statement, equivalent to index-based…
String Methods: split(), strip(), and lower() — Finding unique words involves a pipeline: read file → split lines into words → check membership to avoid duplicates → sort…
String Methods: split(), strip(), and rstrip() — find() returns the starting index of a substring if found, or -1 if not found
String Methods: strip() and rstrip() — The newline character (\n) is a single invisible character that represents the end of a line, despite appearing as two symbols…
String Methods: upper(), lower(), and strip() — Write a program that prompts the user for a filename, opens the file, and prints each line in uppercase using the upper() string…
String Operations and Methods — Sometimes we may want to construct strings from other information. This is where the format() method is useful.
String Operations and Slicing — raw_input() captures text from the keyboard and returns it as a string to your program.
String Replication with the * Operator — The + operator performs concatenation when used with strings, joining them end to end to create a new string.
String Representation of Objects — Special methods enable custom classes to mimic built-in type behaviors through double-underscore naming conventions.
String Slicing — List slicing uses the syntax t[start:stop] to extract a portion of a list, where start is included and stop is excluded.
String Slicing and Indexing — String reversal uses the slice notation [::-1] to read characters backward.
String Slicing and Manipulation in Python — read() loads an entire file into a single string, including all newline characters, and exhausts the file handle resource after…
String Slicing and Substrings — len() is a built-in Python function that counts and returns the total number of characters in a string.
String Slicing Syntax and Boundaries — Use find() to locate a delimiter like a colon within a string, which returns the index of that character.
String Slicing Techniques — A sequence is an ordered set of values identified by integer indices starting at 0; strings are sequences of characters.
String Traversal and Iteration — A slice with start >= end always produces an empty string, not an error.
String-to-Number Conversion in Python — Extract numbers from formatted text lines using a regex pattern with a capturing group and the findall() method.
Strings and String Methods — Strings are objects belonging to the str class and possess methods for checking, searching, and transforming text.
Strings and Text — A string is a sequence of characters -- in practice, strings are basically just the words and text a program works with.
Strings in Python — An example of a literal constant is a number like 5 , 1.23 , or a string like 'This is a string' or "It's a string!" .
Strings: Immutable Character Sequences — Tuples are immutable and therefore do not provide .sort() or .reverse() methods. These methods modify sequences in place, which…
Strong Convexity — Tikhonov regularization augments the loss with λ ‖w‖2.
Structural Pattern Matching (Python 3.10+) — Python has no switch statement; this is a deliberate design choice, not an oversight.
Structural Risk Minimization — Nonuniform learnability is characterized by a countable union of agnostic PAC learnable classes.
Structured Output Learning — A multiclass predictor must combine evidence for several possible classes into one output.
Structured Output Prediction — Ranking problems arrange instances by relevance rather than assigning an isolated label to one instance.
Structured Statements — A block groups multiple statements into one structured statement.
Structuring Code for Readability — As programs grow larger and more complex, they become harder to read due to the density of formal language
Structuring Milestones in AI Vendor Contracts — A successful contract negotiation outcome combined full payment of the outcome milestone, validated metric movement, and…
Subclasses and Inheritance — Object is a special class in package java.lang and is the one class in Java that is not a subclass of any other class.
Subgradient — SGD requires an expectation-based analysis.
Subqueries: Nesting SELECT Statements — SQL uses a single equal sign (=) for equality testing in WHERE clauses, not the double equal sign (==) used in Python.
Subroutine Contracts in Program Design — A precondition is a condition that must be true when a subroutine is called if it is to work correctly, and it is an obligation…
Subroutine Design — A main routine handles the overall flow of a program and hands specific work to subroutines.
Subroutine Parameters — A subroutine parameter lets the same subroutine perform its task on different inputs, with the caller choosing the input on each…
Subroutine Parameters and Arguments — A parameter may have an array type such as int[], so one argument can deliver an entire array to a subroutine.
Subroutine Parameters and Return Values — A parameterized subroutine declares a parameter list in its definition, which specifies the number and types of values a caller…
Subroutine Return Values — A formal parameter is a named slot in a subroutine definition, with a simple identifier as its name and a declared type.
Subroutine Signatures and Parameters — The signature of a subroutine does not include its return type.
Subroutines and Formal Parameters — The scope of a name is the part of the source code where that name is valid, and it is determined by where the declaration…
Subroutines and Modularity — A subroutine is a named group of instructions for a specific task.
Subroutines and Parameters — A functional interface is a Java file that holds a specification for exactly one subroutine, and it is defined in a .java file…
Subsampling layers — Deep convolutional networks organize processing through convolutional, subsampling, and fully connected layers.
Substance Versus Cargo Cult in Consulting Delivery — Exceptions to the document-producing default do exist within the Big Four, usually in specific practices within AI or applied…
Substantial versus Imitative Engagements — The document-versus-code test asks whether an engagement's primary artifact is a document or a deployment.
Subword Tokenization and Vocabulary Design — Byte Pair Encoding is a tokenization method used in natural language processing, including in GPT-2.
Summing Values in a Loop — Data scanning loops follow a three-step construction: initialize variables before the loop, perform computation on each item…
Supervised Instruction Fine-Tuning — Preparing a dataset is a step within supervised instruction fine-tuning.
Supervised Learning — Unsupervised learning finds useful transformations of input data without target values.
Supervised Learning and Hypothesis Classes — PAC-Bayes bounds define a hierarchy over a hypothesis class H.
Support Vector Machines — Early neural networks were explored decades before modern deep learning, but their development was constrained by training…
Support Vector Machines and Margin-Based Learning — Hard-SVM's output w_S is defined by Equation (26.19) and satisfies L_S(w_S) = 0 under the theorem's assumptions.
Support Vector Machines for Large Margin Classification — Margin measures the smallest distance from a training point to a hyperplane.
Surrogate Loss Functions — The Perceptron can be placed inside the Online Gradient Descent framework.
Surveys and Recent Developments in Reinforcement Learning — Choose a resource category according to your reading goal.
SVM Optimization and Regularization — Regularizing the norm of w helps both sample complexity and computation.
SwiGLU Activation Function — GELU is a smooth activation function defined exactly as x times Phi(x), where Phi is the cumulative distribution function of the…
Swing Component Hierarchy — Swing components are arranged in an inheritance hierarchy.
Swing Event-Handling Thread — A Swing Timer's thread only generates events; the handler code that responds to them runs on the Swing event-handling thread, so…
Switch Statements — A switch expression computes and returns a single value, so it is used when the purpose of a switch is to produce a value for an…
Switch Statements with Numeric Values — A switch statement can branch on a String value, letting the program react to a unit name typed by the user.
Symbolic AI — Artificial intelligence is the effort to automate intellectual tasks normally performed by humans.
Symbolic AI and Rule-Based Systems — Classical programming starts with rules written by humans and uses them to process data.
Symmetric Matrices — A symmetric matrix is positive definite when every eigenvalue is positive.
Synaptic Plasticity and Reinforcement Signals — Miller's 1981 Law-of-Effect-like rule included synaptically local contingent eligibility traces.
Syntax and Semantics of Programming Languages — BNF is a notation for writing the syntax rules of programming languages, developed by John Backus and Peter Naur in the late…
Syntax of Java Programs — Syntax is the form a program must take, and the source stresses that knowing syntax is only part of being a successful programmer.
Synthesis Review in the Paving Cycle — The source describes the paving cycle as roughly seven months, built from two three-month stages and one six-week paving stage.