Data Pre-Processing Steps in Deep Learning or Machine Learning :

1. Vectorization :

The process of Labeling Structured or Unstructured Data is called Vectorization.

For Example:

Image of a cat labeled as 1 and Image of Dog as not Cat

2. Normalization :

To avoid extra-computation and waiting extra time we transform large or small values of all useful columns(attributes) between 0 and 1 where their mean must be zero and their standard deviation be 1.

3. Handling Missing values :

Insert reasonable values to missing or NaN values of columns either by mean, median, or mode. you also can use other approaches too.

4. Feature extraction :

Select useful columns/features wisely and split them into labels and input/inputs according to the given problem.