AI Terms 2

Question 1

What is a Knowledge Graph?

Answer 1

Network structure representing relationships between entities:

Stores information as interconnected nodes and edges
Enables semantic search and reasoning
Supports relationship discovery
Used in recommendation systems

Question 2

What is a Graph Database?

Answer 2

Database optimized for storing and querying graph structures:

Native support for relationships
Efficient traversal operations
No need for complex joins
Example: Neo4j

Question 3

What are ACID Transactions?

Answer 3

Properties ensuring database reliability:

Atomicity: All or nothing execution
Consistency: Valid state transitions
Isolation: Transaction independence
Durability: Permanent changes

Question 4

What is Vertex AI Matching Engine?

Answer 4

GCP’s vector database service:

Optimizes similarity search
Scales to billions of vectors
Low-latency retrieval
Integrates with Vertex AI

Question 5

What is Reinforcement Learning?

Answer 5

Learning through environment interaction:

Uses rewards/penalties
Learns optimal actions
Explores vs exploits
Example: Game AI

Question 6

What is Supervised Learning?

Answer 6

Learning from labeled data:

Input-output pairs
Classification/regression
Requires labeled datasets
Example: Spam detection

Question 7

What is Unsupervised Learning?

Answer 7

Learning patterns without labels:

Clustering
Dimensionality reduction
Pattern discovery
Example: Customer segmentation

Question 8

What is Continuous Evaluation?

Answer 8

Ongoing model performance monitoring:

Real-time metrics
Performance degradation detection
Automated testing
Quality assurance

Question 9

What is Data Drift?

Answer 9

Changes in input data distribution:

Feature value shifts
Input pattern changes
Requires monitoring
May trigger retraining

Question 10

What is Model Drift?

Answer 10

Degradation of model performance:

Relationship changes
Accuracy decline
Required retraining indicators
Performance monitoring

Question 11

What are Model Parameters?

Answer 11

Learned variables during training:

Weights and biases
Adjusted automatically
Learned from data
Define model behavior

Question 12

What is Learning Rate?

Answer 12

Step size for model updates:

Controls convergence speed
Affects training stability
Hyperparameter
Typically 0.001-0.1

Question 13

What is Batch Size?

Answer 13

Training samples per iteration:

Affects memory usage
Impacts training speed
Trade-off with accuracy
Hyperparameter

Question 14

What are Batch Workloads?

Answer 14

Processing large data volumes:

Non-real-time
Resource-intensive
Scheduled processing
Bulk operations

Question 15

What are Real-time Workloads?

Answer 15

Immediate data processing:

Low latency
Stream processing
Immediate responses
Online serving

Question 16

What is Deterministic Processing?

Answer 16

Fixed outcome for given input:

Predictable results
No randomness
Reproducible
Rule-based

Question 17

What is Predictive Analysis?

Answer 17

Forecasting future outcomes:

Based on patterns
Probability-based
Uses historical data
What might happen

Question 18

What is Prescriptive Analysis?

Answer 18

Recommending actions:

Suggests solutions
Optimization-based
What should be done
Action-oriented

Question 19

What are Cold Start Problems?

Answer 19

Initial recommendation challenges:

New users/items
No historical data
Limited predictions
Requires special handling

Question 20

What does “Scales Quadratically” mean?

Answer 20

Resource needs grow with square of input:

O(n²) complexity
Exponential resource growth
Performance consideration
Common in attention mechanisms

Question 21

What is Neural Architecture Search?

Answer 21

Automated network design:

Optimizes model structure
AutoML component
Searches architecture space
Automated optimization

Question 22

What is Contextualization in AI?

Answer 22

Process of understanding input in its full context:

Considers surrounding information
Improves accuracy
Enhances relevance
Critical for NLP tasks

Question 23

What is Automated Parameter Adjustment?

Answer 23

Automatic optimization of model parameters:

During training process
Uses gradient descent
Optimizes weights/biases
No manual intervention

Question 24

What is Manual Hyperparameter Optimization?

Answer 24

Human-guided tuning of model settings:

Trial and error
Expert knowledge
Grid/random search
Performance tracking

Question 25

What is Entity Recognition?

Answer 25

Identifying named entities in text:

Names, locations, dates
Organization names
Custom entity types
Part of NLP pipeline

Question 26

What is Content Classification?

Answer 26

Categorizing content into predefined groups:

Document categorization
Topic labeling
Automated sorting
Pattern recognition

Question 27

What is Syntax Analysis?

Answer 27

Understanding grammatical structure:

Sentence parsing
Grammar checking
Language understanding
Structure identification

Question 28

What is Unstructured Text Classification?

Answer 28

Categorizing free-form text:

No predefined format
Natural language
Pattern extraction
Automated categorization

Question 29

What are Matrices in ML?

Answer 29

2D arrays of numbers:

Data representation
Linear transformations
Feature organization
Mathematical operations

Question 30

What is Augmented Decision Making?

Answer 30

AI-assisted human decisions:

AI recommendations
Human final choice
Enhanced insights
Decision support

Question 31

What is Automated Decision Making?

Answer 31

Full AI decision automation:

No human intervention
Rule-based decisions
Model-driven choices
Automated actions

Question 32

What are Rollback Points?

Answer 32

Safe states to revert to:

Version control
Recovery points
Safety mechanism
Change management

Question 33

What is an Array and how is it used in programming and ML?

Answer 33

A data structure that stores ordered elements:
Core Definition

Fixed or dynamic-size collection
Sequential memory storage
Zero-based indexing (usually)
Same-type elements (in most languages)

Types

One-dimensional

Simple list: [1, 2, 3, 4]
Single row of elements
Linear access pattern

Two-dimensional

Matrix-like structure: [[1,2], [3,4]]
Rows and columns
Grid representation

Multi-dimensional

Nested structures
Tensor representation
Complex data organization

Common Operations

Insertion/deletion
Random access by index
Iteration/looping
Slicing/subsetting

ML Applications

Dataset storage
Feature matrices
Mini-batch organization
Image representation (2D/3D arrays)

Performance

Fast random access: O(1)
Insertion/deletion: O(n)
Memory efficient
Cache-friendly