Neural_Networks PPT FOR CLASS 12 AI CBSE .pptx

NEURAL NETWORK PPT FOR CLASS 12 AI CBSE

1. Neural Networks - Simplified Explanation Class 12 Artificial Intelligence (CBSE)

2. What is a Neural Network? • An AI model inspired by the human brain • Made of interconnected 'neurons' (nodes) • Each neuron processes inputs and passes information • Learns patterns and makes predictions • Example: weather prediction,Facial Recognition

3. Neural Network Diagram

4. Key Components of a Neural Network • Input Layer: Receives data (features) • Hidden Layers: Learn complex patterns • Output Layer: Produces final result • Weights & Biases: Adjust strength of connections • Activation Function: Adds non-linearity • Loss Function: Measures error • Optimizer: Adjusts weights to reduce error

5. Working of a Neural Network 1. Forward Pass: • Multiply inputs by weights, add bias • Apply activation function • Pass to next layer 2. Loss Calculation: • Compare predicted output with actual 3. Backward Pass: • Calculate error contribution of each weight • Adjust weights using optimizer 4. Repeat training until accuracy improves

6. Neural Network Diagram

7. Main Components of a Neural Network Neurons (Nodes) A neuron is a basic processing unit of a neural network (similar to a nerve cell in the brain). Each neuron receives input, processes it, and passes an output to the next layer. Every neuron performs a simple computation: Output=Activation( Weights×Inputs+Bias )Output = Activation(Weights \times Inputs + Bias)Output=Activation( Weights×Inputs+Bias ) 🧩 Example for students: Think of each neuron as a student who reads information (input), thinks (processes), and gives an answer (output).

8. Main Components of a Neural Network Layers Neural networks are made up of different layers of neurons: Input Layer: Receives the data (e.g., height, weight, marks). Hidden Layer(s): Process the inputs by applying weights, bias, and activation. Output Layer: Produces the final prediction (e.g., “Pass” or “Fail”, “Cat” or “Dog”). 🧩 Analogy: Input layer = Eyes 👀 (collect data) Hidden layer = Brain 🧠 (process data) Output layer = Mouth 🗣️ (gives decision)

9. Main Components of a Neural Network Weights Weights determine how important each input is to the output. Every connection between two neurons has a weight value . These are adjusted during training so that the network makes accurate predictions. 🧩 Example: In predicting exam results, “Study Hours” may have a higher weight than “ Favorite Color .”

10. Main Components of a Neural Network Bias A bias is an additional parameter that helps adjust the output along with the weighted inputs. It allows the model to shift its decision boundary and learn more flexibly. 🧩 Analogy: If you always add some fixed marks to everyone’s score (grace marks 🎓), that’s bias.

11. Main Components of a Neural Network Connections These are the links between neurons of one layer to the next. Each connection carries the input multiplied by its weight. Together, they form a network of information flow. 🧩 Think of it like: Roads connecting cities — data travels through these roads to reach the final destination

12. Main Components of a Neural Network Activation Function It decides whether a neuron should activate or not . Converts raw input values into meaningful outputs (e.g., 0 or 1). Common activation functions: Sigmoid – outputs between 0 and 1 ReLU (Rectified Linear Unit) – outputs 0 if input < 0, else the same number Tanh – outputs between -1 and

13. Main Components of a Neural Network Learning Rule (Algorithm) It defines how the network adjusts weights and biases based on the error. The goal is to reduce the difference between predicted output and actual output. Example: Gradient Descent adjusts weights to minimize loss. 🧩 Analogy: Learning rule = Teacher correcting students after a test (you learn from mistakes).

14. Main Components of a Neural Network Propagation Functions Two main propagation processes occur in a neural network: a. Forward Propagation Data moves from input to output through layers. Each neuron processes data and passes it forward. The network predicts the output.

15. Main Components of a Neural Network b. Backward Propagation After prediction, the model compares its output with the correct answer. It calculates error (difference) and moves backward to adjust weights using the learning rule. This helps the model “learn.” 🧩 Analogy: Forward = Writing an exam Backward = Checking the paper and learning from mistakes

16. Main Components of a Neural Network Component Description Analogy Neuron Basic processing unit Brain cell Weight Importance of input Trust in advice Bias Adjustable constant Grace marks Layers Stages of processing Eye–Brain–Mouth Activation Function Decision gate On/Off switch Learning Rule Adjusts weights Teacher’s feedback Forward Propagation Passes data forward Writing exam Backward Propagation Learns from errors Checking answer sheet

17. Neural Network Diagram

18. Applications of Neural Networks • Image and speech recognition • Predictive text and chatbots • Self-driving cars • Healthcare diagnostics • Fraud detection • Personalized recommendations (Netflix, Amazon)

19. Recap • Neural networks are inspired by the brain • They learn by adjusting weights and biases • Key parts: Input, Hidden, Output layers • Use activation, loss, optimizer to train • Applied in many industries around us