1. Understand the problem you are trying to solve:
Prepare a well-defined problem, instead of a vague description, and be straight to the point of what you are trying to achieve in your project.
If you have experience working in data science projects make sure you know
Who is your client?
What exactly is your client asking you to solve? (His Business Problem)
2. How did you collect the data?
Being a Data person there is data involved obviously in your work
Make sure you know from what kind of resources and the pipeline to collect the data
3. What exactly is in my data? “Data Understanding”
Once data got collected, one needs to understand the information contained in it within a high level.
- Be strong with all the methods which are used to understand the data.
4. Data Processing:
After gaining a better understanding of data from the previous step, you have identified below abnormalities in the data
- Corrupt records
- Missing values
- Outliers and other challenges you will have to manage
First, you need to clean the data to convert it to a form that you can further analyze or apply some machine learning algorithms.
5. Application of Machine Learning or Statistical Models:
Be thorough with the model pipeline from start to end with all the algorithms you have worked with including optimization.
6. Communicate the Results of Your Analysis or Deployment:
All the analysis and technical results that you come up with are unless if you cannot explain to your stakeholders what they mean.