Neurosciences Developmental Neural Network Role of Nervous System

Neuroscience's Developmental & Neural Network • Role of Nervous System: Signal Processing Stimulation 1 Organ Nervous System Brain 2 • Signal forward Method of Neuron Using the Action Potential 1

• Neuroscience's Developmental * Understanding of Brain Activity: Finding Identity of Man * Important Advances in Recent Neuroscience ◎ New Nerve Signaling Substances or Controlled Substances Discovery ex) Neuropeptides ◎ Cytophysiology → Sensory and Exercise Behaviors Related ex) Peptide Hormones in Brain * Development of Cognitive Science Complex Behavior Base: Circuits of Brain [In many Parts of Brain Compositive Involved] * Magnetic Resonance Imaging(MRI) Technology ⇒ Shorten Time Between Stimulus and Reaction [2 s] 2

• Computer and Brain * Application of Neural Networks ◎ Happening in each Part of Nervous System ⇒ Simulation ◎ Useful Device Production in Industrial ◎ General Hypothesis Testing to Brain Activity * Difference between Computer and Brain(B) Action Potentials Generation Rate Computer(C) Brain < Computer Operating methods Parallel Series Input / Output Multiple Input, Multiple Output Minority Input, Minority Output Remember How Action Potentials [Encoding(X)] Exactly pulses form of 0 and 1 [Encoding(O), Accuracy] 3

• Computer and Brain * Difference between Computer and Brain(B) Computer(C) Output for Same Input Variable Same Memory Storage Space Activity Perform Space = Storage Space Activity Perform Space ≠ Storage Space Design How Natural selection (each other for generations) Elaborate Design of Engineers → Biological Evolution * Computer with Brain Comparison → Justification minutely Parallel Operation of Brain Sequential Operation Mechanism 4

• Developmental Process of Neural Networks * Learning about ◎ Educator(X), Learning by Repetition[To only Simple Problems Applies] ex) Hopfield Model ⇒ connection Strength Adjustment: Hebb Method ◎ Educator(O), Learning by Teaching ex) Interactive Activation Model ⇒ connection Strength Adjustment: Hebb Method ◎ Competitive Learning - Idea: Behavior of Neural Networks → Winner-Take-All Algorithms [Winner-Take-All mechanism: Largest Output of Output Layer Units, all Units Inhibitory] ◎ Adequacy of Algorithm Biological Algorithms < Non-biological Algorithms ◎ Generality & Strong Biological Algorithms < Non-biological Algorithms 5

* Neural System Identification Close Design in Actual Neural Networks + Well known Backpropagation of Errors How to minimize errors Claim Close Results to General Property of Actual Neural Network 6