Practical Machine Learning: Part 1
This is a series of articles to help other Data Scientists based on my learning
“How do I improve my model?”
This is the most common question I get asked by other data scientists. This question, to me, is the equivalent of “What should I do now?”
There is no one right answer to this question. Just like for the latter question, I would start answering the former with “It depends.”
The truth is, it really depends on multiple considerations, for example:
- What do you actually mean by accuracy here?
- Why is improving accuracy important? Are you sacrificing something else just to improve accuracy?
- What does improving accuracy mean to your end user?
- Can you afford to add more training data to improve accuracy?
In this post, I wanted to provide a logical Decision Tree to help you decide on what step to take to improve your model accuracy. I will limit the Part 1 to just the ML models.