📘 Lesson · Lesson 90

Activation Functions

What is an Activation Function?

💡 At a Glance

An activation function decides whether a neuron "fires". It adds non-linearity so neural networks can learn complex patterns.

Common Functions

Function	Output Range	Use
Sigmoid	0 to 1	binary output / probabilities
Tanh	-1 to 1	hidden layers (zero-centered)
ReLU	0 to infinity	most common in deep networks

In Code

Python

import numpy as np
def relu(x): return np.maximum(0, x)
def sigmoid(x): return 1 / (1 + np.exp(-x))

print(relu(-3), relu(5))      # 0 5
print(round(sigmoid(0), 2))   # 0.5

0 5 0.5

Summary

Activation functions add non-linearity to neural networks.
ReLU is most common; Sigmoid for probabilities; Tanh is zero-centered.

Activation Function क्या है?

💡 एक नज़र में

Activation function तय करती है कि neuron "fire" करे या नहीं। यह non-linearity जोड़ती है ताकि neural networks जटिल patterns सीख सकें।

Common Functions

Function	Output Range	Use
Sigmoid	0 से 1	binary output / probabilities
Tanh	-1 से 1	hidden layers (zero-centered)
ReLU	0 से अनंत	deep networks में सबसे common

Code में

Python

import numpy as np
def relu(x): return np.maximum(0, x)
def sigmoid(x): return 1 / (1 + np.exp(-x))

print(relu(-3), relu(5))      # 0 5
print(round(sigmoid(0), 2))   # 0.5

0 5 0.5

सारांश

Activation functions neural networks में non-linearity जोड़ती हैं।
ReLU सबसे common; Sigmoid probabilities के लिए; Tanh zero-centered।

← Back to Python Tutorial