Date your Data

The ultimate guide to know how to take the best of her out !

This article intends to illustrate the data science project cycles steps with analogy of story of a boy who’s trying to get into a serious relationship with a lady .

So let’s discover how he is willing to do that…

First, let me introduce you to the story actors :

• Kim : The data scientist
• Lina : The data

Find your Date

This step may seem obvious, however, it is not !

In real world cases, extracting the sufficient amount of information ,like data, can either lead you to the right or the wrong direction.

The better data’s quality is obtained , the less uncertainty we would have , and vice versa .

So it’s important to use the right type of data, and to keep in mind that data collection may be challenging and time consuming.

Plan the Date

• What are you going to be doing with your data ?
• Set a goal to reach at the end of your study in order to avoid going through misleading roads (note that it happens frequently ).

Get to know your Date

• Do some visualisations and summary statistics.
• Understand what your data is about . Know each feature well and ask of its of quality .
• Apply the EDA (exploratory data analysis) technic.

Look for anything weird

What we mean by weird, is a data with an unusual value & it could be a good or bad thing .

Some outliers can give you better understanding of your data , while others can be ignored … If any , investigate why the values are missing in your data and find the best way to impute if possible.

Take your date out

• Try out various machine learning models and train it on your data .
• Each model is designed for certain types of data .Find one and optimize it.

Evaluate your date

• Determine your metrics and see if all of your training was worth it .
• Does it meet your expectation ? If not, you’re going to do more training or get more data.

Conclusion :

Those were the key steps that every data scientist should follow to provide a reliable data interpretation and insightful results .

If you have any questions or comments feel free to leave your feedback below or you can always reach me on LinkedIn.

Thank you for reading and have a good date .. AKA Data ;)