CertNexus CET flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

CIA triad

The three core security objectives: confidentiality (only authorized access), integrity (data is accurate and unaltered), and availability (systems and data are accessible when needed).

Symmetric encryption

An encryption method where the same secret key is used to encrypt and decrypt data. It is fast and efficient but requires secure key distribution between parties.

Asymmetric encryption

An encryption method using a public key to encrypt and a mathematically related private key to decrypt. It solves key distribution but is slower than symmetric encryption.

Multi-factor authentication

Authentication requiring two or more distinct evidence types, such as something you know, something you have, or something you are. It reduces risk from stolen passwords.

Defense in depth

A layered security strategy that applies multiple overlapping controls (physical, network, endpoint, application, data, identity) so failure of one layer does not compromise the whole.

Incident response phases

The NIST incident response lifecycle: preparation; detection and analysis; containment, eradication, and recovery; and post-incident activity including lessons learned.

Zero Trust

A security model that assumes no implicit trust based on network location. It verifies every request explicitly, enforces least privilege, and assumes breach.

IoT architecture layers

Typical IoT stack: perception/device layer with sensors and actuators, network connectivity layer, edge/fog processing layer, cloud/application layer, and business layer.

MQTT protocol

A lightweight publish/subscribe messaging protocol over TCP using a broker. It is widely used in IoT for low-bandwidth, high-latency, or unreliable networks.

CoAP protocol

A REST-like request/response protocol over UDP designed for constrained IoT devices. It supports low overhead and is secured with DTLS.

LoRaWAN

A low-power wide-area network protocol for long-range, low-bandwidth IoT communication. It suits battery-operated sensors sending small infrequent messages.

Edge computing

Processing data near the source (at gateways or devices) instead of only in a central cloud. It reduces latency, bandwidth use, and dependence on connectivity.

IoT OTA updates

Over-the-air firmware updates delivered to IoT devices. Secure OTA requires signed images, integrity verification, rollback protection, and authenticated delivery channels.

Secure boot

A hardware-rooted startup process that verifies firmware signatures before execution. It prevents unauthorized or tampered code from running on IoT devices.

Supervised learning

A machine learning approach trained on labeled input-output pairs to predict outcomes for new data. Common tasks are classification and regression.

Unsupervised learning

A machine learning approach that finds patterns in unlabeled data. Common techniques include clustering, association rules, and dimensionality reduction.

Reinforcement learning

A machine learning paradigm where an agent learns by interacting with an environment, taking actions, and receiving rewards or penalties to maximize cumulative reward.

Precision and recall

Precision = TP/(TP+FP), measuring how many predicted positives are correct. Recall = TP/(TP+FN), measuring how many actual positives were found.

Overfitting

A modeling error where a model fits training data noise and performs poorly on new data. Mitigations include more data, regularization, dropout, and cross-validation.

Bias-variance tradeoff

The balance between a model too simple (high bias, underfitting) and too complex (high variance, overfitting). Total error combines bias squared, variance, and irreducible error.

Generative AI

AI systems that create new content such as text, images, or code, typically using large language models, GANs, or diffusion models. Risks include hallucination and data leakage.

Responsible AI

A framework for developing and deploying AI ethically, covering fairness, accountability, transparency, privacy, safety, robustness, and meaningful human oversight.

Data science lifecycle

The end-to-end process: problem definition, data acquisition, cleaning, exploratory analysis, modeling, evaluation, deployment, and ongoing monitoring.

Descriptive vs inferential statistics

Descriptive statistics summarize observed data (mean, median, standard deviation). Inferential statistics use sample data to draw conclusions about a larger population.

Correlation versus causation

Correlation measures linear association between variables. Causation means one variable produces change in another. Confounders can create spurious correlations.

Type I and Type II errors

Type I error is a false positive: rejecting a true null hypothesis. Type II error is a false negative: failing to reject a false null hypothesis.

Data normalization

Scaling numerical features to a common range, such as 0 to 1, or standardizing to zero mean and unit variance. It improves model convergence and comparability.

Sampling bias

A distortion where the sample is not representative of the population, such as survivorship, selection, or response bias. It threatens the validity of conclusions.

GDPR

The EU General Data Protection Regulation. It requires a lawful basis for processing, grants data subject rights, and mandates breach notification within 72 hours.

Data minimization

A privacy principle requiring collection and processing of only the personal data necessary for a specified purpose. It limits exposure and regulatory risk.

Anonymization vs pseudonymization

Anonymization irreversibly removes identifiers so data cannot be linked to individuals. Pseudonymization replaces identifiers but allows re-identification with additional data.

Algorithmic bias

Systematic unfairness in automated decisions caused by biased training data, proxy variables, sampling errors, or feedback loops. It requires auditing and mitigation.