What is the role of activation function in Neural Networks ? The role of the activation function in a neural network is to produce a non-linear decision boundary via non-linear combinations of the weighted inputs. A neural network classifier is essentially a logistic regression classifier without the hidden layers. The non-linearity to a neural network is added by the hidden layers using a sigmoid or similar activation functions.
Quite often, I take the Uber Pool ride to my office in the morning hours of Bangalore's heavy traffic. Although I get a bit disappointed every time a ride request comes to the driver (I prefer to take the sit beside the driver) but given that pooling is more economical and the thought that I am helping Bangalore traffic, makes me feel better. But I do get frustrated, when the […]
Expectation Maximization is a quite an old tool/concept in the Machine Learning domain. Although it is an old tool but it took me quite some time to grasp the concept and the intuition behind it given that most tutorials and articles out there explain it with heavy mathematical equations. But eventually I found out that, the maths behind the intuition is pretty simple to understand, only the long equations might […]