14. Neural Networks

Neural Networks

---

Topic: Neural Networks

Course: GMLC

Date: 24 March 2019  

Professor: Not specified

---

Resources

---

• https://developers.google.com/machine-learning/crash-course/introduction-to-neural-networks/video-lecture

• https://developers.google.com/machine-learning/crash-course/introduction-to-neural-networks/anatomy

• https://developers.google.com/machine-learning/crash-course/introduction-to-neural-networks/playground-exercises

• https://developers.google.com/machine-learning/crash-course/introduction-to-neural-networks/programming-exercise

Key Points

---

• Neural Networks are used to solve non-linear problems by adding hidden layers & activation functions

• Hidden layer

  • Sum of all inputs

• Activation function

  • Sum of all hidden layers ran through an activation function (weighted sum of input values)

  • Rectified Linear Unit

  • Sigmoid Function

  • σ(w⋅x+b)

• Structure

  • Set of nodes in layers

  • Set of weights representing the connection

  • Set of biases

  • An activation function

• Neuron

  • Node which takes multiple inputs and generates a single output value using activation functions

Check your understanding

---

• Know the structure of Neural Networks

• Understand what are activation functions & what are they used for

Summary of Notes

---

• Neural Networks are used to solve non linear problems using hidden layers summing the weights on inputs ran through an activation function