Supervised Learning - Classification Week 9 Challenge

Answer 1:

Neural network are a subset of machine learning and are at the heart of deep learning algorithms. Their name and structure are inspired by the human brain, same way that biological neurons signal to one another.

Neural networks rely on training data to learn and improve their accuracy over time.

Answer 2:

Deep Learning is a subset of Machine Learning which is essentially a neural network with three or more layers.This network helps in predicting the target value.

Answer 3:

Activation function is used in neural network to introduce non-linearity in the given dataset by adding a bais to the evaluating function. This helps in estimating result for more complicated problem and even dealing with noisy data efficiently.

Answer 4:

Different types of activation function are :

1. Sigmoid : It is most common activation function which can take values from 0 to 1 and is montonically increasing. `1/(1+e^-x)`
2. Tan-h : It tends to centre the output to 0 which has an effect of better learning on the subsequent layers. `(e^x - e^-x)/(e^x + e^-x)`
3. Relu : It is a simple non-linear activation function that has the advantage of removing the problem of vanishing gradient.

Answer 5:

Backpropagation is the technique used by neural network to learn from training dataset and hence improve performance by tuning the weights of neural network.Proper tuning of the weights allows you to reduce error rates and make the model reliable by increasing its generalization.