Bias - Bigger Model

Course Q&A Deep Learning Specialization Structuring Machine Learning Projects
1

Hi,

One of the methods recommended to avoid Bias is training a “Bigger Model” what do we exactly mean by bigger?

2

Hi, @Moutasem_Akkad!

The most common way to make a model “bigger” is either adding hidden layers (depth) or increasing the neurons / filters in the linear / convolution layers (width). Figure 2 in Mingxing & Quoc shows a great example of this.

1 Like
3

Hello,
Yes, adding more layers and more neurons to each layer makes the model bigger and this helps to avoid bias