Seven Steps to Hands-On Machine Learning

1. Collecting data
2. Prepare data
3. Select a model
4. Training
5. Assessment
6. Parameter adjustment
7. Prediction

Let's use an example to see how each step works.

Case: Distinguishing between wine and beer

Collecting data

We bought a bunch of different types of beer and red wine at the supermarket, marked all the wines with their color and alcohol content, and recorded the data.This step is very important because the quantity and quality of the data directly determines the quality of the predictive model.

Data Preparation

In this example, our data is very neat, but in practice, the data we collect will have many problems, so it will involve data cleaning and other work.
When there is no problem with the data itself, we divide the data into 3 parts:

• Training set (60%): used to train the model
• Validation Set (20%): Make sure the model is not overfit
• Test set (20%): used to evaluate model performance

Select a model

Researchers and data scientists have created many models over the years.Some are great for image data, some are great for sequences (such as text or music), some are for numerical data, and some are for text-based data.
In our case, since we only have 2 features, color and alcohol, we can use a small linear model, which is a fairly simple one.

Training

Most people think this is the most important part, but it's not the case~ The quantity and quality of data, and the choice of model are more important than the training itself (3 minutes on the training knowledge desk, more important is10 years in the audience).
This process does not require human participation, the machine can complete it independently, and the whole process is like doing an arithmetic problem.Because the essence of machine learning is the process of transforming problems into mathematical problems and then solving them.

Assessment

Once the training is complete, you can evaluate whether the model is useful.This is where our previously reserved validation and test sets come into play.The evaluation indicators mainly include precision rate, recall rate, and F value.
This process allows us to see how the model is making predictions about numbers we haven't seen yet.This is meant to represent how the model will perform in the real world.

Parameter adjustment

After the evaluation is complete, we can further improve the training by adjusting the parameters.When we train, we implicitly assume some parameters that we can adjust to make the model perform better.

Forecast

Our 6 steps above are all for this step.This is also the value of machine learning.At this time, when we buy a new bottle of wine, just tell the machine its color and alcohol content, and he will tell you that the beer is still red wine.

References

https://easyai.tech/ai-knowledge-hub/