Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
37 Cards in this Set
- Front
- Back
|
Tasks studied by Tenenbaum using bayesain probabilities
|
concept learning, causality, and property induction
|
|
|
3 characteristics of Grammar (pinker):
|
productive, symbolic, and combinatorial
|
|
|
Occam's razor principle in the context of Bayesian inference
|
More complex models have less probability. Therefore, the preference is toward simple models.
|
|
|
what makes a certain code more or less codable?
|
a single word is more codable then a phrase
|
|
|
what are two important lessons w/ regards to web-scale learning?
|
-Large quantities of unlabeled data
-Memorization is a good policy |
|
|
Prescriptive Grammar
|
specifies rules for correct language use (rules)
|
|
|
Generative Grammar
|
specifies rules followed in producing sentences (production from the rules)
|
|
|
2 factors which determine the internal structure of the Language faculty
|
- Genetic endowment
- External linguistic data |
|
|
What does the Language of Thought Hypothesis explain?
|
the combinatorial properties of thought. (A^B-->AB)
|
|
|
Finite state machine w/ infinite time and tape to make calculations
|
Turing Machine
|
|
|
Chomsky's Minimalist Theory
|
Operation of merge which combines 2 expressions together
|
|
|
Finite state machine
|
machine w/ finite number of symbols, instructions, and states of memory.
|
|
|
Poverty of Stimulus
|
input of data required to learn a language underdetermines the output.
|
|
|
elements which are required of representational machinery to learn language
|
words and word combinations
|
|
|
Deterministic Program
|
It has no choice, actions are determined by the input and state of the memory.
|
|
|
3 things needed for a symbol system
|
- set of symbols
- Things for those symbols to refer too - set of rules |
|
|
what is the best model for the world
|
the world!
|
|
|
Design features of the human language according to Jackendoff
|
open learnt vocabulary
concentration of symbols Inflectional morphology (change form of word to give it extra meaning) |
|
|
Design features unique and w/in human communication (slides)
|
signal, user, message, and rules
|
|
|
What is categorization
|
Agent+environment+interaction=
Categorization! |
|
|
categorization is cognition
|
categorization determines how we interact w/ the environment (can be seen as an algorithm)
|
|
|
categorization is an algorithm for:
|
- interaction
- Knowledge representation - Learning -generalizations and abstractions |
|
|
language and categorization
|
allow categories to be conceptualized and acquired though hearsay.
|
|
|
what is categorization and why is it important?
|
reveals the input conditions of information when being transfered from the environment to the agent
|
|
|
Cognition seen autopoiesis
|
perhaps cognition is a form of complex bodily function (self creation/organization)s
|
|
|
Faultless disagreements and relativism
|
under relativism faultless disagreements exist. Truth is relative to one's perspective.
|
|
|
Faultless disagreement
|
when 2 ppl believe contradicting statements yet neither is at fault
|
|
|
when do faultless disagreements arise?
|
- value judgments
- probability - belief |
|
|
Equivalence Schema
|
It is true that p iff p
|
|
|
Connectionism
|
interconnected network of simple and uniform units
|
|
|
Ecological view
|
only way to make an intelligent machine is to ground it in the world
|
|
|
NP-hard
|
number of steps to check an answer grows as exponentially
|
|
|
Behaviorism
|
there is no difference between actions and mental process
|
|
|
Turing machine
|
theoretically can compute every every computable function
|
|
|
affordences
|
things represented in the environment by the mind
|
|
|
supervienence
|
non-causal dependencies. for example, appearance facts supervene on chemical facts
|
|
|
functionalism
|
mental states constitutive by functional role
|
