Concepts starting with H

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

Habit Learning and Goal-Directed Behavior — Goal-directed and habitual behavior rely on different, though not necessarily exclusive, brain contributions.
Habitual and Goal-Directed Processes in the Brain — Model-free and model-based reinforcement learning provide a distinction for studying brain processes.
Habitual Behavior — The experiment changes reward value after learning and measures whether behavior adjusts.
Halfspace Hypotheses — The same dot product ⟨w, x⟩ can be interpreted probabilistically by logistic regression or deterministically by a halfspace.
Halfspace Learning — Surrogate losses replace other losses for purposes such as computational convenience.
Halfspaces — The base class B provides the individual hypotheses used as building blocks.
Hallucination in Language Models — Temperature controls how random a model's output is. Higher values let lower-probability tokens surface more often, but the…
Hand-Calculating Layer Connections — The source's network has two inputs, two outputs, and two hidden layers of 30 and 20 nodes, and its code reports 752 trainable…
Handling API Rate Limits and Throttling — SQLite serves as a local cache to store geocoding results, preventing redundant API calls to the same locations.
Handling Binary Data and Encoding in Python — A socket connection to port 80 establishes a reliable communication channel with a web server through a three-way handshake.
Handling Collisions in Hash Tables — A hash table stores key/value pairs in an array and uses the key to find where a pair is stored, avoiding a scan through…
Handling Concurrent Inserts and Database Locking — Auto-generated primary keys shift the responsibility for ID management from the application to the database, eliminating manual…
Handling Customer Disagreements — Ask questions about the customer's needs before you propose or reject a solution, because the answers decide which design fits.
Handling Customer Objections Without Escalation — The customer simulation tests composure under pressure more than any other quality, especially when the right answer is hard to…
Handling Database Errors and Exceptions — A database connection is established using sqlite3.connect(filename), which creates a link to a local database file or remote…
Handling Duplicate Data with INSERT IGNORE — A multi-table relational schema organizes data across separate tables to eliminate redundancy and improve data integrity. The…
Handling Edge Cases in Linked List Insertion — Insertion into a sorted linked list must place the new node where its value belongs, which is often in the middle of the list…
Handling Errors Gracefully in Programs — A program is correct when it accomplishes the task it was designed to perform, and this is judged against the inputs it is meant…
Handling Errors in Web Scraping — BeautifulSoup's soup('a') method finds all anchor tags in parsed HTML; tag.get('href', None) safely extracts the href attribute…
Handling Errors with try and except — The input() function returns a string, so you must use int() to convert it to an integer for arithmetic operations.
Handling Events in Java GUI Programs — A JCheckBox has two states, selected and unselected, and the state is represented by a boolean that is true when selected and…
Handling Exceptions in File Operations — A File object represents a file name, not a file. The named file may or may not exist, and creating the object creates nothing…
Handling Exceptions in Python — The sys module provides system-specific functionality and acts as an interface between Python and the operating system.
Handling Exceptions with try..catch — An exception is an event that disrupts the normal flow of control in a program.
Handling File Errors and Exceptions — Writing to a file requires opening it with mode 'w', calling write() to insert data, and closing the file to flush the buffer to…
Handling Foreign Key Relationships — INSERT OR IGNORE silently skips an insertion if it would violate a constraint, allowing your program to continue without an error.
Handling Invalid Directories — A Java File object represents a file or directory, and creating one with new File(name) asks the program to refer to a path…
Handling IT Pushback in High-Priority Projects — The customer's IT and security teams are expected to push back on Full Disk Encryption and related parts of the proposal.
Handling JSON Parsing Errors and Exceptions — json.loads() transforms a JSON string into a Python data structure by parsing the text and building corresponding lists…
Handling Large Files Efficiently — read() loads an entire file into a single string, including all newline characters, and exhausts the file handle resource after…
Handling Model Outputs Safely — A language model, by itself, does not differentiate between user-provided input and its own generation, so any separation of…
Handling Mouse Events in GUI Programming — A GUI program has no main() routine that runs step by step from beginning to end. It must respond to events that happen at…
Handling Multiple Exception Types — An ungraceful exit occurs when an unhandled exception causes a program to crash and print a traceback instead of terminating in…
Handling Nested XML Structures — findall retrieves a Python list of all XML nodes matching a specified path, enabling batch processing of similar structures.
Handling Network Errors and Timeouts — Buffered reading prevents memory exhaustion by reading large files in fixed-size blocks and writing each block to disk before…
Handling Personal Health Information in AI Deployments — HIPAA governs personal health information in the United States, and its obligations apply when an AI system touches PHI.
Handling Unknown Questions in Interviews — An interview tests how you think, not how much you have memorized, so an unknown question is an opportunity to show reasoning.
Handling Unknown Words in Tokenization — BPE begins its vocabulary with every individual single character, so any text built from those characters can be represented.
Handling User Input in Swing Components — JTextField and JTextArea are both subclasses of the abstract class JTextComponent, which provides their shared behavior such as…
Handling Whitespace in Strings — startswith() checks if a string begins with a specific prefix and returns True or False—it does not search for the substring…
Handling Window Events — The default close operation, set with setDefaultCloseOperation, decides what happens when the user clicks a JFrame's close box.
Handling XML Parsing Errors — fromstring converts an XML string into a tree structure, enabling programmatic access to its data.
Handoff Between Solution and Product Teams — The engagement documentation is a structured record, not a slide deck, and it exists so that a team inheriting the work can…
Handoff Process Overview — The handoff process transfers day-to-day operation of the system from the FDE to the customer's internal team at the end of an…
Handoffs Between Engagement Teams and Product Teams — A structured record of an engagement answers ten questions covering the underlying problem, customer priorities and…
Hard-Fail Criteria in Automated Evaluation — LLM-judge calibration checks a judge's grades against a set of handlabeled contracts so that trust is based on measured…
Hard-Margin SVM Optimization — Soft-SVM preserves the margin constraint while allowing controlled violations.
Hard-SVM — Margin measures the smallest distance from a training point to a hyperplane.
Hardness of Learning 3-term DNF Formulae — Representation independent learning separates the original hypothesis class from the class used to produce the output.
Hardware Considerations for AI Models — Inference efficiency matters because a model's usability depends heavily on its inference cost and latency.
Hash Codes and Hash Functions — A hash table stores key/value pairs in an array and uses the key to find where a pair is stored, avoiding a scan through…
Hash Functions and Unordered Maps — Creating a new hash function by combining existing hash functions is a simple approach that the source describes as often very…
Hash Tables and Collision Handling — hashCode() is defined in the Object class, returns an int, and every Java object has one.
Headers and Namespaces in the C++ Standard Library — Real C++ programs are rarely written in the bare language; libraries make most tasks far simpler to express.
Heaps and Heap Implementation of Priority Queues — A priority queue is an abstract data type whose items each carry a priority, and whose remove operation returns an item with the…
Heuristic Search Methods — Samuel's checkers player combined a search role with a learning role.
Hidden States in Neural Networks — An encoder-decoder architecture divides translation between an encoder that reads the whole input and a decoder that writes the…
Hiding Superclass Members in Java — super is a special variable used inside a subclass to refer to the object's superclass view, and it remembers only that the…
Hierarchical Clustering — Linkage-based clustering repeatedly merges clusters, beginning with single-point clusters.
High-Variance Bets in Vertical AI Startups — The decision to join an AI-native startup depends on three factors together: the vertical, the team, and the stage.
Higher-Order Functions — A lambda statement creates an anonymous function object with one parameter and one expression.
Higher-Order Polynomial Bases — Higher-order bases offer greater approximation capability for complicated functions.
Hilbert Space — H2 ◦ S consists of all m-coordinate inner-product vectors produced by allowable w vectors.
Hinge Loss — Sample complexity asks how many training examples are needed to control unseen-data performance.
Historical Context and Advanced Topics — The curse of dimensionality is associated with high-dimensional data.
Historical Context and Influences on Reinforcement Learning — Reinforcement learning concerns learning through interactions between an agent and its environment.
Historical Foundations of Actor-Critic Architectures — Miller's 1981 Law-of-Effect-like rule included synaptically local contingent eligibility traces.
History of C++ Standardization — Growing use of C++ made formal standardization necessary, and the source places the recognition of that need in 1987.
Holder's Inequality — The H1 proof reduces a Rademacher-complexity calculation to a finite-set problem.
Homogeneous and Non-homogeneous Linear Separators — VC dimension is a measure of classification capacity based on shattering.
Homogeneous Halfspaces — The compressed object is a subset of training points, not the weight vector itself.
Homogeneous Linear Separators — The Natarajan dimension is a complexity measure for a hypothesis class.
How AI Models Generate Text Token by Token — AI models are probabilistic because they sample responses from a set of possible answers, so the same prompt can yield different…
How Attention Weights Are Computed — Attention mechanisms exist to address a specific need in neural networks, so the reasons for using them come before the…
How Computers Execute Instructions — A compiler translates an entire program from high-level source code to machine language in one complete pass, then saves the…
How Computers Execute Programs — The CPU is the Central Processing Unit.
How Computers Follow Instructions — Humans and computers have complementary cognitive strengths, not a universal ranking: humans excel at reading and understanding…
How Computers Store and Process Information — Programming is a creative partnership where humans provide ideas, creativity, and judgment while computers provide speed…
How Computers Understand Instructions — Programming is a problem-solving skill that develops through practice and feedback, not a collection of syntax rules to memorize.
How Foundation Models Generate Output — Autoregressive models generate tokens sequentially, so output time accumulates with the number of tokens.
How Language Models Learn from Text — Large training datasets for GPT- and BERT-like models are diverse, comprehensive text corpora measured in billions of words.
How Large Language Models Are Pretrained — A token is a unit of text that a model reads, and dataset sizes are commonly reported as token counts, which roughly track the…
How Large Language Models Generate Text — The transformer is the architecture that LLMs use to process their input.
How Loops Change Execution Flow — Code file order (top to bottom) is not the same as execution order (the actual sequence the program runs).
How Monte Carlo Tree Search Works — MCTS uses four stages: Selection, Expansion, Simulation, and Backpropagation.
How Neural Networks Learn: Loss Functions and Optimization — Training deep neural networks with many layers can be challenging, and the source names vanishing or exploding gradients as a…
How Neural Networks Process Data — The three fields form a nested hierarchy: artificial intelligence contains machine learning, which contains deep learning.
How Objects Communicate: Calling Methods and Passing Data — A zone is a self-contained unit combining code (methods) and data (attributes) with well-defined interactions with other zones…
How Programs Execute: From Code to Action — Input devices capture user actions and send data into the computer; output devices receive data from the computer and present it…
How Python Dictionaries Work Internally — Hashability is the property that allows an object to be used as a dictionary key. A hashable object must produce the same hash…
How Python Executes Code — However if you place this code in a Python script and this error occurs, your script immediately stops in its tracks with a…
How the CPU Executes Instructions — Main memory is fast and nearly as quick as the CPU, but it is volatile and loses all data when power is cut off.
How to Import Modules in Python — How It Works First, we import the sys module using the import statement. Basically, this translates to us telling Python…
How to Parse XML — XML (eXtensible Markup Language) is a text-based format for encoding structured, hierarchical data using custom tags that…
How to Practice Programming Effectively — Programming mastery, like language fluency, requires time and patience to absorb and internalize concepts.
How to Read and Interpret Error Messages — Python requires precise, exact syntax because it is a tool without emotions or judgment. Even small mistakes cause Python to…
How to Read and Understand Error Messages — but that is much uglier and error-prone. Second, the conversion to string would be done automatically by the format method…
How to Read Error Messages — Syntax errors break Python's grammatical rules and prevent the program from running. Python reports them immediately with a line…
How to Write a Function — Functions allow you to name groups of statements, which is the foundation of readable code.
HTML Structure and Parsing — Web scraping is a program-based technique for retrieving web pages and extracting specific data from their HTML source code…
HTTP and Web Communication Protocols — An API is a published contract between applications that specifies the exact rules for requesting and receiving services.
HTTP Protocol Basics — A socket is a two-way connection between two programs that allows both reading and writing of data, similar to a file but…
HTTP Protocol Fundamentals — curl and wget are command-line tools for retrieving files from the web on Unix-like systems using HTTP or FTP protocols.
HTTP Request and Response Structure — recv() returns the amount of data currently available in the network buffer, which may be less than the requested amount and…
HTTP Request Headers and Query Parameters — API keys are the mechanism vendors use to identify users and monitor their service consumption.
HTTP Request Structure and Headers — Binary files must be accumulated in a bytes buffer and cannot be printed as they arrive like text files.
HTTP Requests and Response Objects — Binary files require byte-for-byte copying from a remote URL to your local disk.
HTTP Requests and Responses — Web scraping is a program-based technique for retrieving web pages and extracting specific data from their HTML source code…
HTTP Requests and Web Clients — BeautifulSoup is a Python library designed to parse real-world HTML, which is often malformed in ways that cause strict XML…
HTTP Status Codes and Error Handling — HTTP responses follow a three-part structure: status line, headers, blank line, and body.
HTTP: The Protocol Behind Web Communication — An API is an application-to-application contract that publishes rules for accessing services provided by one program for use by…
HTTPS and Encryption in Transit — Vendors require cryptographically signed messages to verify both the source of requests and that they have not been altered in…
HTTPS and Secure Communication — API keys can be included in requests via two methods: as a URL query parameter or as part of POST data.
HTTPS and Web Security Basics — HTTP is a standardized, text-based protocol that defines how web clients and servers communicate using a simple request-response…
Human Evaluation of LLMs — Conversational performance is the ability of LLMs to engage in human-like communication by understanding context, nuance, and…
Hyperparameter Optimization — Normalization makes samples more similar by centering and scaling their values, which can support learning and generalization.
Hyperparameters and Training Configuration — A compute-optimal model delivers the best performance achievable for a fixed compute budget, so teams usually start from the…
Hyperscaler and lab versions of the FDE motion — Hyperscaler-lab convergence is the pattern in which the hyperscaler and lab versions of the FDE motion move toward each other…
Hypothesis — Classifier error measures the probability of an incorrect prediction on a random data point.
Hypothesis Class — A hypothesis class H is a set of functions from X to Y.
Hypothesis Class Complexity — The Perceptron predicts online from a weight vector and an incoming input vector.
Hypothesis Classes — Finite classes contain a limited number of hypotheses, often because their representations are bounded.
Hypothesis Classes and Learning Algorithms — Weak learnability allows a learning algorithm to be useful even when its accuracy is only slightly better than random guessing.
Hypothesis Classes and Training Error — Compression in the unrealizable setting seeks minimum error within H, not necessarily zero error.
Hypothesis Classes and VC-Dimension — Approximation error measures the gap between the best hypothesis in a class and the true hypothesis.
Hypothesis Classes in Machine Learning — Shattering is an all-labelings requirement, not a one-labeling test.
Hypothesis Classes of Neural Networks — Neural networks with an appropriate depth-2 structure can represent every Boolean function.
Hypothesis Selection — Overfitting is a mismatch between excellent training performance and poor true-distribution performance.