- Spending Less Time in Understanding the data and EDA.
- Not Communicating well.
- Spending more time on Theory without practical application.
- Focusing on Accuracy over Understanding how the model works.
- Giving Preference to Tools over Business problems.
- Ignoring Outliers.
- Using L1, L2 Regularization without Standardization.
- No proper understanding on how to transform categorical variables.
- Not picking the right loss function.
- Not Focusing on the Distribution of data.
- Correlation Does Not Imply Causation.
- Assuming the Algorithms are more important then Domain Knowledge.
