• Mistakes can be hard to avoid sometimes

• Through challenges and mistakes one can discover new data trends and discover more data patterns.

• However, there is a small margin for error when working on data science projects.

Let’s talk about:

and how to avoid them!

Mistake # 1: Low Data Quality and Volume

• can make it difficult to create predictive models

Potential Solutions:

• Simulate the data- must be done carefully

• Buy data sets (purchase-intent data)

Mistake # 2: Starting to work without proper exploration

• It’s not enough to have intuition or theories about your data set

Solution:

• Exploratory Analysis!

Mistake # 3: Having high expectations

• Can’t just fling data at a computer algorithm and expect it to all work out

• As a newbie data scientist, focus on small projects and quick wins

Mistake # 4: Avoiding a control group to test a new data model

• Without a control group, you can’t measure how good the model is working

Mistake # 5: Not updating your data model

• Although a model is solving a business problem at present, that doesn’t mean it is an all time solution

Mistake # 6: Avoiding team members

• Make sure you are on the same boat with business

Mistake # 7: Avoiding simple tools

• Avoid wasting resources if the task can be executed using a simple method

Mistake # 8: Constant use of implementations that misalign with your problem

• There’s a high availability of open source implementation of a lot of algorithms, but might not align with the problem you are trying to solve!

Solution: - You can build your own version for a specific use case

The End