Models Of Memory Information Processing And Neural Network
Memory is explained through two major models: the Information Processing Model (which treats memory like a computer system with stages of input, storage, and retrieval) and Neural Network Models (which view memory as patterns of activation across interconnected neurons). Both highlight different aspects of how the brain encodes, stores, and recalls information.
Information Processing Model -
- Analogy: Memory works like a computer system.
- Stages:
– Encoding: Input of information through sensory channels.
– Storage: Retaining information in short-term or long-term memory.
– Retrieval: Accessing stored information when needed. - Key Components:
– Sensory Memory: Brief storage of sensory input.
– Short-Term/Working Memory: Temporary holding and manipulation of information.
– Long-Term Memory: Permanent storage of knowledge and experiences. - Strengths: Explains sequential flow of information and highlights capacity limits (e.g., STM’s 7 ± 2 items).
- Limitations: Oversimplifies memory as linear; doesn’t fully capture emotional or associative influences.
Neural Network Models -
- Analogy: Memory is distributed across networks of neurons.
- Principles:
– Neurons and synapses form circuits that encode experiences.
– Memory is stored as patterns of activation rather than discrete “files.”
– Learning involves synaptic plasticity* (strengthening or weakening of connections). - Examples of Models:
– Hopfield Network: Explains associative memory (recalling whole patterns from partial cues).
– Continuous Attractor Networks: Model working memory and sustained neural activity.
– Reservoir Networks: Capture dynamic processing and flexible recall. - Strengths: Reflects biological plausibility, explains parallel processing and resilience to damage.
- Limitations: Complex to model; harder to map directly to conscious experiences.