Getting familiar with Machine Learning | Neural networks vocabulary

Neural networks always performs well with training.

To perform training on the neural network we need data.

Following is the example data set of iris flowers.

It has 5 columns.

Sepel length, Sepel width , Petal Length , Petal Width decides the species of the  iris flowers.

We are going to predict the species of a particular flower depending on these characteristics.

According to that knowledge we have 4 characteristics of data which we know.   We call them features.   

So  Sepel length, Sepel width , Petal Length , Petal Width are the features of this data set.

Species field is the label which we are going to predict.

Rows of this data set are called examples.

Which means , example is a set of features and labels for a particular record.

We can optimize the neural network model by training.

When we train the model by using the examples with labels we call it supervised machine learning.

When we train the model by using the examples without labels we call it unsupervised machine learning.

In this data set we use we have data for the label in the examples.

If we train our neural network using this data set it is supervised  machine learning.

If we do  not have data for the labels and we train the model using that data it is unsupervised machine learning.