Hello there,
I am new to participating in machine learning competitions and have been learning a lot through the challenges hosted on this platform. Although; I have encountered some difficulties when dealing with data quality issues and would appreciate some guidance from more experienced members.
I have noticed that the datasets sometimes include missing values; inconsistent data formats; or even outliers that seem to skew the models performance. While I understand that part of the challenge is to handle such data effectivel;, I find myself struggling to determine the best practices for cleaning and preprocessing these datasets.
What are the most effective strategies for dealing with missing data? Is imputation generally preferred, or are there cases where it is better to exclude certain data points altogether?
How do you decide when to normalize or standardize data? Are there specific indicators in the dataset that suggest one method over the other? 
What approaches do you use to identify outliers in a dataset? Once identified, how do you decide whether to remove, modify, or keep these outliers in the dataset?
Also; I have gone through this post; https://www.talend.com/resources/machine-learning-data-quality-mlops/ which definitely helped me out a lot.
What are some efficient ways to address issues like varying date formats or inconsistent categorical values in a dataset?
Thank you in advance for your help and assistance. 