1
$\begingroup$

I've followed lots of tutorials on Machine Learning but in each of these, they go for a different strategy so it's quite confusing for me. I want to Know that what are the operations involved and what are the correct ordering of these.

AS of now, I think the process and the ordering are ->

  1. Get Data
  2. Delete Duplicates
  3. Find Missing Values and Outliers
  4. Create New Features
  5. Deal with missing values and Outliers
  6. Build a base model
  7. Find the best features to select
  8. Try and find different Models
  9. Select the BEST model
  10. Hypertuning of the Model

Please Do Provide if something is missing and correct the sequence.

$\endgroup$
1

1 Answer 1

1
$\begingroup$

This is a broad topic and not every problem has the same flow. And, it's rather a cycle. The first step is correct, you need to gather data. Then, you perform exploratory data analysis. This may include but not limited to data ingestion, cleaning, transformation etc. And you don't always delete duplicates. This step may also overlap with feature generation. Some problems (e.g. computer vision) might not need feature generation by the way; the features are also learnt from raw data via the first layers of a DNN.

A typical approach is to build a baseline model to compare your later improvements. These improvements might include additional feature generation or trying new methods. While trying new methods, you try to do your best with the method at hand. So, hyperparameter optimization (HPO) is also a part of this process if applicable. Therefore, Your number 10, i.e. HPO stage, comes before selecting the best model. You might still want to tweak your selected model at later stages by entering another HPO/feature selection loop. From an application perspective, deployment of your model is the last stage.

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.