core technologies Flashcards
how many industry eras? describe the 4th.
5? debatable, but 4 for sure
4th- 2000s
= ongoing transformation of industries and society through the integration of advanced digital technologies
blurring lines, merge with human lives
a lot of innovation
risk: worse inequality
demand for skills that ppl do not have
components: big data, blockchain, it, ai and machine learning, cloud computing
what is cloud computing?
= remote storage that allows access from any device, data synchronisation and backup and collaboration and access
without cloud computing
- high upfront costs
- complex management
- scalability issues
- risks of obsolescence
- disaster recovery
with cloud computing
- no hardware purchase
- pay-as-you-go
- reduced management complexity
- built-in disaster recovery
- rapid deployment
types of cloud- pros and cons
public:
pro- third-party provider, available to anyone, scales quickly and convenient
con- lack of customisation, governance issues
private:
pro- resources dedicated to one business, more privacy for sensitive data
con- expensive
hybrid:
pro- greater flexibility
con- more complex
implications of cloud computing
business operation:
- cost efficiency
- rapid deployment
- greater scalability and flexibility
- global reach
innovation and entrepreneurship:
- lower barriers to entry
- new business model
- increased competition
environmental:
- energy efficiency
- resource optimisation
challenges and risks:
- data privacy and protection
- cross-border data transfer
- security risks
4 Vs of big data
Variety- various data
Volume- a lot of data
Velocity- speed of data generation and process
Veracity- uncertainty or quality of data
types of data
structured- 20%- less storage, easy to manage
unstructured- 80%- more storage, diff to manage, img, video, etc- can be sometimes converted into structured
data analytics for structured data
descriptive- what happened?
diagnostic- why it happened?
predictive- what will happen?
prescriptive- what should we do?
data analysis for unstructured data
text analysis- summarise, sentiment analysis
audio analysis- speech analysis. musical info retrieval, voice identification, environmental management, health care, security
implications data analytics
business:
- data-driven decision making
- new business models based on real-time data analytics
challenges and risks:
- data privacy and protection
- poor data quality
- bias and fairness in analytics
- ethical use of data
AI and machine learning
structured data, pre-processing data needed
ai= teaching computers to think and make decisions like humans
machine learning (90s)= train computers to learn from examples- patterns learned from data
rule-based systems
= type of AI that applies sets of logical rules to data or input to make decisions, perform actions or generate outputs
- rely on pre-defined rules- if-else- dictate the processing of inputs and determine the system’s behaviour
cons:
- dependency od domain knowledge
- not flexible enough
- lack of learning capabilities
Deep learning revolution and AI
unstructured data, pre-processing data not needed
- 2010s- by how the human brain works- teach computers to learn from large amount of data using ‘neural networks’ that mimick the brain structures
- 2020s- GPT-3- milestone for Natural Language Processing and Generative AI
traditional vs generative AI
traditional- recognise patterns or make decisions based on already existing data
generative- generate new data
