Basics
Define the PDP Approach.
A theory suggesting that cognitive processes occur through networks of connected units that work simultaneously.
What is activation?
The level of activity that a unit has at a given moment
Symbolic Approach
A traditional cognitive approach that suggests thinking happens through step-by-step symbol manipulation.
When you recognize a friend’s face quickly by processing many visual features at once.
parallel processing in face recognition
What are the units in the PDP Approach?
The small processing elements in a PDP network that represent information
distributed representation is?
A pattern of activation across multiple units that represents information
What are patterns of activation across units?
The PDP approach suggests knowledge is stored in these patterns rather than single locations
A system that learns to recognize handwritten letters after seeing many examples
PDP learning model for pattern recognition
T/F: The small processing elements in a PDP network that represent information are units.
True
When activation spreads from one unit to connected units throughout the network is labeled as..
spreading activation
T/F: Is neural networks in the brain the system that inspired the structure of PDP models.
What is distributed representation in word recognition?
Understanding a word even if one letter is missing because other units help fill the gap
What is learning through weight changes?
The process of increasing or decreasing connection strength based on experience
T/F: The value that determines how weakly one unit influences another is connection weight
False: keyword "strongly"
What is simultaneously, in parallel?
A key difference between PDP and symbolic models is that PDP processes information in this way
T/F: Learning language patterns through repeated exposure instead of memorizing implicit rules is connectionist learning.
False: it is through explicit rules.
What is parallel processing?
The idea that many units process information at the same time rather than one step at a time
The process of adjusting connection weights to improve performance after experience or feedbacks is learning in a ___ network
neural
The concept that damage to part of a network leads to gradual performance loss rather than complete failure is what?
Graceful degradation
What are artificial neural networks in AI?
Modern technology inspired by PDP models that learns from large datasets to recognize patterns.