Computer Science
Modified: 2020-02-24
- Relevant Areas of Computer Science
- AI
- Definition
- The attempt to make computers perform intelligent actions.
- Subfields of AI
- Natural Language Processing
- The attempt to get computers to understand spoken
speech.
- Chess and other games
- Games are an old area in AI
- Checkers and Backgammon are understood
- Chess is falling fast (Deep Blue)
- Robotics
- "Hard AI"
- Real world constraints
- Expert Systems
- Another success of AI
- Consist of knowledge base and inference engine
- Knowledge Retrieval and Data Mining
- Newer area
- Consists of finding difficult to see patterns in large
groups of data (i.e., stock market)
- Agents
- Consist of software that can search a database
- Many use the Internet as their domain
- Knowledge Representation
- How is knowledge represented?
- How do humans represent knowledge?
- How should computers represent knowledge?
- Parallel Distributed Processing
- Alternative to von Neumann serial design
- Difficult to implement the software
- Machine Vision
- Another difficult area
- Problem is how to analyze and interpret the large amount of
data that is collected
- Machine Learning
- How can machines be made to learn for themselves (i.e.,
like children do)?
- Communication
- Transmission Control Protocol (TCP) – used for data transmission
- User Datagram Protocol (UDP) – used by programs to send short datagram messages
- Internet Control Message Protocol (ICMP) – messages used for diagnostic or error-generating purposes
- Hypertext Transfer Protocol (HTTP) – application protocol that uses hyperlinks between nodes containing text
- Post Office Protocol (POP) – used by local email clients to retrieve email from a remote server over TCP IP
- File Transfer Protocol (FTP) – protocol to transfer computer files from a server to a client and vice versa
- Internet Message Access Protocol (IMAP) – a communication protocol used by email clients to retrieve messages from a mail server over TCP IP
- Human Computer Interaction (HCI)
- Usability/User experience
- Complex problem
- Desktop metaphor
- Social media
- Videoconferencing
- Ubiquitous computing
- Kasparov and AI
- 23 years after losing to Deep Blue (CS class watched live!)
- Blessing (but first knowledge worker to be displaced by a machine, he says)
- Only 4% of US jobs require creativity, he says
- Radiology: rather have a nurse than a big MD star (challenge the machine)
- Humans as shepherds to AI (nudging them)
- Does not believe in AGI (artificial general intelligence)
- Humans are evil, not machines (discussing ethical AI)
- AlphaZero and Magnus Carlsen
Back to Main Page