Intro to AI 2 final terms (Lecture Neural Networks)

Perceptron

A basic unit of a neural network that makes binary decisions using weights and bias.

Key Parts of a Perceptron:

• Inputs: Features of the data (e.g., X1, X2)

• Weights: Values that determine the importance of each input

• Bias: A constant added to the result to shift the decision boundary.

• Activation Function: Usually a step function that outputs 1 if the total input is above a threshold, and 0 otherwise.

• Output: The prediction (1 or 0)

Activation Function

Function that determines if a neuron fires (e.g., threshold, sigmoid).

Bias Term

A constant added to the weighted sum to shift the decision boundary.

Weight

A value that determines the importance of a given input in a neuron.

Linearly Separable

A dataset that can be separated by a straight line.

XOR Problem

A logic function that is not linearly separable; can't be solved by a single perceptron.

Multilayer Perceptron

A neural network with one or more hidden layers, capable of solving non-linear problems.

McCulloch-Pitts Neuron

An early model of a neuron using binary input, summation, and threshold activation.

The activity of the neuron is an ‘all-or-none’ process (either it fires - output 1, or it doesn’t - output 0)

Threshold Function

Fires if the weighted input sum exceeds a threshold.

Learning Rate

Controls how much weights are adjusted during training.

Biological Neurons

The basic building blocks of the nervous system, which process and transmit information through electrical and chemical signals. Biological neurons consist of a cell body, dendrites, and an axon.

Action Potential

A brief change in the electrical potential across the membrane. The inside of the neuron becomes positively charged.