model parameters VS hyper parameters

In machine learning, there are a lot of theory and key words that can get confusing at times. A common example is the confusion that happens between what a model parameters and hyper parameter is.

Model parameter are variables that are scoped to the model itself, they are needed for the model to make accurate predictions. The values of these parameters can be learned from the data we feed the model. Model parameters values are learned during the training.

Hyper parameter on the other hand are like variables used for configuring the learning algorithm not the model. for example how many layers or hidden neurons do you want in your neural network, the learning rate. Hyper parameter are manually set.