Question

What are some of the shortcomings of traditional ‘one-hot’ vector encodings’? What are some of the shortcomings of traditional 'one-hot' vector encodings'?

a. They tend to be very short
b. They tend to contain many zero elements
c. They tend to be very dense
d. They tend to be relatively sparse

Answer 1

The shortcomings of traditional 'one-hot' vector encodings include: containing many zero elements, being very dense, and relatively sparse. These encodings can be inefficient in terms of space and computational resources.

Step-by-step explanation:

The shortcomings of traditional 'one-hot' vector encodings include:

1. They tend to contain many zero elements
2. They tend to be very dense
3. They tend to be relatively sparse

'One-hot' vector encodings are commonly used in machine learning and natural language processing tasks. They represent categorical variables as binary feature vectors, where each element corresponds to a unique category. However, these encodings can be inefficient in terms of space and computational resources, especially when dealing with large datasets.