DSST MIS flashcards

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

Shannon capacity formula

Maximum error-free data rate C = B × log2(1 + S/N), where B is bandwidth in hertz and S/N is the linear signal-to-noise power ratio. Doubling bandwidth or raising SNR increases capacity.

Circuit versus packet switching

Circuit switching reserves a dedicated end-to-end path for the entire session. Packet switching divides data into packets that share links dynamically, giving better utilization for bursty traffic at the cost of variable delay.

OSI model layers

Seven layers from bottom: Physical, Data Link, Network, Transport, Session, Presentation, Application. Each layer provides services to the layer above and communicates logically with its peer layer on the remote host.

CSMA/CD versus CSMA/CA

CSMA/CD detects collisions after they occur and is used on wired Ethernet. CSMA/CA avoids collisions using reservations and acknowledgments before sending, which is necessary on wireless media where collision detection is impractical.

Subnet usable host count

For a subnet with h host bits, usable addresses = 2^h − 2, subtracting the network and broadcast addresses. A /26 mask leaves 6 host bits, yielding 62 usable hosts.

Attenuation and crosstalk

Attenuation is the weakening of a signal as it travels over distance, measured in decibels. Crosstalk is unwanted signal coupling from one conductor or channel into an adjacent one.

Symmetric versus asymmetric encryption

Symmetric encryption uses one shared secret key and is fast, suited to bulk data. Asymmetric encryption uses a public/private key pair, solving key distribution and enabling signatures, but is computationally much slower.

Digital signature mechanics

The sender hashes the message and encrypts the hash with the sender's private key. The receiver decrypts with the sender's public key and compares hashes, providing authentication, integrity, and nonrepudiation.

CIA triad

Confidentiality keeps data hidden from unauthorized parties, integrity ensures data is unaltered and trustworthy, and availability ensures systems and data are accessible when needed. Security controls are classified against these three goals.

IDS versus IPS

An intrusion detection system monitors traffic and raises alerts but does not block. An intrusion prevention system sits inline and can drop or reset malicious traffic in real time, trading some latency for active defense.

Stateful packet inspection

A stateful firewall tracks the state of active connections and permits return traffic belonging to an established session. A stateless packet filter examines each packet independently against header rules only.

Risk in security terms

Risk is the potential for loss, commonly modeled as threat × vulnerability × impact. A vulnerability is a weakness; a threat is an actor or event that could exploit it; exposure is what could be lost.

DFD balancing rule

When a process is decomposed into a child data flow diagram, the child's external inputs and outputs must exactly match those of the parent process. Unbalanced diagrams indicate inconsistent analysis.

Waterfall versus iterative SDLC

Waterfall completes each phase sequentially with requirements frozen early. Iterative and agile approaches repeat analysis, design, and build in short cycles, incorporating feedback and allowing requirements to evolve.

Critical path in PERT/CPM

The critical path is the longest-duration path through a project network and determines the shortest possible completion time. Activities on it have zero slack, so any delay there delays the whole project.

Entity-relationship cardinality

Cardinality states how many instances of one entity relate to another: one-to-one, one-to-many, or many-to-many. Many-to-many relationships become associative tables when mapped to relational tables.

Feasibility study types

Feasibility assesses whether a system should be built: technical (can it be built), economic (cost versus benefit), operational (will users adopt it), schedule (can it be delivered in time), and legal.

Joint application design

JAD is a structured workshop bringing users, managers, and developers together to elicit and resolve requirements in concentrated sessions. It shortens requirements gathering and builds consensus compared with serial interviews.

Expected monetary value

EMV = sum over outcomes of (probability × payoff) for a decision alternative. Under risk, the alternative with the highest EMV is preferred; EMV does not capture risk tolerance or worst-case loss.

Maximin and maximax criteria

Maximin chooses the alternative whose worst outcome is best, reflecting pessimism or risk aversion. Maximax chooses the alternative with the best possible outcome, reflecting optimism. Minimax regret minimizes maximum opportunity loss.

Break-even quantity formula

Break-even quantity = fixed costs ÷ (unit price − unit variable cost). The denominator is contribution margin per unit; higher margin or lower fixed cost reduces the volume needed to cover costs.

Net present value rule

NPV discounts all expected future cash flows to present value at a chosen rate and subtracts the initial investment. A positive NPV means the project adds value; among competing projects, higher NPV is preferred.

Structured versus unstructured decisions

Structured decisions are routine and rule-based, suited to automation. Unstructured decisions are novel and ambiguous, requiring judgment. Semi-structured decisions mix both and are typically supported by decision support systems.

Tacit versus explicit knowledge

Explicit knowledge is codified, documented, and easily shared, such as manuals and databases. Tacit knowledge is personal, experiential, and hard to articulate, transferred best through mentoring and shared practice.

SECI knowledge conversion model

Four modes: socialization (tacit to tacit via shared experience), externalization (tacit to explicit via articulation), combination (explicit to explicit via synthesis), and internalization (explicit to tacit via learning by doing).

Communities of practice

Informal groups of practitioners sharing a domain of expertise who interact regularly to solve problems and share techniques. They are especially effective for transferring tacit knowledge that formal documentation cannot capture.

Data-information-knowledge hierarchy

Data are raw unprocessed facts, information is data organized with context and meaning, and knowledge is understanding applied to decisions and action. Each level adds interpretation and increases value.

ETL versus ELT

ETL extracts source data, transforms it on a separate server, then loads it into the warehouse. ELT loads raw data first and transforms it inside the target system, leveraging warehouse compute power.

Star versus snowflake schema

A star schema has one fact table joined to denormalized dimension tables, giving simpler queries. A snowflake schema normalizes dimensions into related subdimensions, reducing redundancy but adding joins and complexity.

Slowly changing dimension types

Type 1 overwrites the old value and loses history. Type 2 inserts a new row with version and effective dates, preserving full history. Type 3 adds a previous-value column to hold limited history.

Fact table grain

Grain specifies what a single fact table row represents, such as one line item per sale. Grain must be declared before selecting dimensions and measures, because it determines which analysis is possible.

Association rule lift

Lift = confidence(A→B) ÷ support(B). Lift above 1 means A and B co-occur more than chance, below 1 means less than chance, and equal to 1 means they are statistically independent.

Apriori principle

If an itemset is frequent, all its subsets are frequent; if an itemset is infrequent, all its supersets are infrequent. This property prunes the candidate search space in association rule mining.

Supervised versus unsupervised learning

Supervised learning trains on labeled examples to predict a class or value, covering classification and regression. Unsupervised learning finds patterns in unlabeled data, covering clustering and association rules.

Precision and recall

Precision = true positives ÷ (true positives + false positives), the share of predicted positives that are correct. Recall = true positives ÷ (true positives + false negatives), the share of actual positives found.

Overfitting in data mining

Overfitting is when a model captures noise specific to the training data and performs poorly on new data. It is detected with holdout or cross-validation and reduced by pruning, regularization, or more data.