Modern computing rests on four deep pillars:
- AI — systems that learn, reason, and generate.
- Cybersecurity — defending systems, identities, and data.
- Networking — moving information efficiently and securely.
- Programming — the languages, paradigms, and abstractions that make software possible.
This A–Z glossary gives you a high‑precision vocabulary across all four domains, designed for advanced learners and professionals.
A
- Adversarial_Attack — Inputs crafted to fool AI models into misclassification.
- Asymmetric_Cryptography — Encryption using public/private key pairs.
- API_Gateway — A control layer managing and securing API traffic.
B
- Backpropagation — The gradient‑based algorithm that trains neural networks.
- BGP — The routing protocol that directs global internet traffic.
- Buffer_Overflow — A memory‑corruption vulnerability enabling code execution.
C
- Containerization — Packaging applications with dependencies for isolated execution.
- CORS — Browser security rules controlling cross‑origin requests.
- Cipher_Suite — A set of algorithms defining how TLS encrypts connections.
D
- Diffusion_Model — A generative AI architecture that denoises data to create images or text.
- DDoS — Attacks that overwhelm systems with massive traffic.
- DNSSEC — Cryptographic protection for DNS integrity.
E
- Embedding — A vector representation capturing semantic meaning.
- Endpoint_Detection_and_Response — Real‑time monitoring and threat response on devices.
- Edge_Router — A router that connects internal networks to external providers.
F
- Federated_Learning — Training AI models across distributed devices without centralizing data.
- Fuzzing — Automated vulnerability discovery using random or malformed inputs.
- Frame_Relay — A legacy WAN protocol for packet‑switched networks.
G
- Gradient_Descent — Optimization method minimizing model loss.
- GRC — Governance, Risk, and Compliance frameworks in cybersecurity.
- gRPC — A high‑performance RPC framework using HTTP/2.
H
- Homomorphic_Encryption — Encryption enabling computation on encrypted data.
- Honeypot — A decoy system designed to attract attackers.
- HSTS — A policy forcing HTTPS to prevent downgrade attacks.
I
- Inference_Engine — The runtime system executing trained AI models.
- IAM — Identity and Access Management for controlling user permissions.
- IPSec — A suite of protocols securing IP communications.
J
- JIT_Compiler — A runtime compiler optimizing code execution.
- JWT — A token format for stateless authentication.
- Jumbo_Frames — Large Ethernet frames improving throughput.
K
- K‑Means — A clustering algorithm grouping similar data points.
- Kerberos — A network authentication protocol using tickets.
- Kubernetes_Operator — A controller automating complex app lifecycle tasks.
L
- LLM — Large Language Models trained on massive text corpora.
- Load_Shedding — Dropping low‑priority traffic to preserve system stability.
- Linter — A tool that analyzes code for errors and style issues.
M
- Model_Distillation — Compressing a large model into a smaller one.
- MITM_Attack — Intercepting communication between two parties.
- MPLS — A high‑performance routing technique using labels.
N
- Neural_Architecture_Search — Automated design of neural network structures.
- NAC — Controls that restrict device access to networks.
- Namespace — A container for identifiers preventing naming conflicts.
O
- Overfitting — When a model memorizes training data instead of generalizing.
- OAuth2 — A protocol for delegated authorization.
- Observability — The ability to understand system state from logs, metrics, and traces.
P
- Prompt_Engineering — Designing inputs to optimize LLM outputs.
- Penetration_Testing — Ethical hacking to identify vulnerabilities.
- P4_Programming — A language for programming network data planes.
Q
- Quantization — Reducing model precision to improve performance.
- QoS_Policies — Rules prioritizing network traffic types.
- Query_Optimizer — A database component that chooses efficient execution plans.
R
- Reinforcement_Learning — Training agents through rewards and penalties.
- Red_Teaming — Simulated adversarial operations to test defenses.
- Routing_Table — A map of paths used by routers to forward packets.
S
- Self_Supervised_Learning — Models learning from unlabeled data using internal signals.
- SIEM — Systems that aggregate and analyze security logs.
- SDN — Software‑defined networking separating control and data planes.
T
- Transformer — The architecture powering modern LLMs.
- TLS_Handshake — The negotiation process establishing encrypted sessions.
- Type_Inference — Automatic deduction of variable types in programming.
U
- Unsupervised_Learning — Discovering patterns in unlabeled data.
- UEBA — Analytics detecting anomalous user behavior.
- UDP_Hole_Punching — A NAT traversal technique for peer‑to‑peer networking.
V
- Vector_Database — Databases optimized for high‑dimensional embeddings.
- Vulnerability_Scanning — Automated detection of security weaknesses.
- VXLAN — A protocol for scalable virtualized networks.
W
- Weight_Sharing — Reusing parameters across model layers to reduce size.
- WAF — A firewall protecting web applications from attacks.
- WPA3 — The latest Wi‑Fi security standard.
X
- XGBoost — A high‑performance gradient‑boosting algorithm.
- XOR_Cipher — A simple symmetric encryption method using XOR.
- XDP — A high‑speed Linux packet‑processing framework.
Y
- YARA_Rules — Pattern‑matching rules for malware detection.
- YAML_Pipelines — Declarative CI/CD workflows defined in YAML.
- Yield_Keyword — A Python construct for creating generators.
Z
- Zero_Shot_Learning — AI models performing tasks without explicit training examples.
- Zero_Day — A vulnerability unknown to the vendor.
- ZKP — A cryptographic method proving knowledge without revealing it.
