Page 63 - Reclaim YOUR DIGITAL GOLD (without audio)

Data ColleCtion Harvesting

is not considered, one may end up adjusting the model’s
parameters for an extended period of time.

These parameters are commonly referred to as “hyper-
parameters.” Tuning or modifying these hyperparame-
ters is still regarded as an art form and an experimental
process. This procedure has a significant impact on the
dataset, model, and training procedure.
We can finally put the model to work and see how it
performs in real-world problems after we have completed
the evaluation process and are satisfied with the training
and hyperparameter settings.

To provide answers to questions, machine learning
obviously relies on data. As a result, we get to answer
some questions during the inference stage, also known
as prediction. This is the culmination of our efforts, the
point at which the benefits of machine learning become
apparent.
Our model can now correctly identify whether the
beverage is wine or beer based on the color and
percentage of alcohol in it.

This is the most basic aspect of artificial intelligence or
machine learning, and it all begins with data collection;
without data, the entire structure becomes inoperable.

43