[TBD]Learning from data science experts

Roles in data science

• Data engineer

  • Work on data engineering

• Machine learning engineer

  • Actively involved in deploying ML models

• Business intelligence

  • Involved in visualization of ML output via Tabeleau etc for business work

Data science manager role

• Balance between release and experimentation

  • Data scientists sometime try to drill too much deep.

  • Need to balance with the release need

Tips for success

• Ensure you are doing everything right

  • Data can't lie you. You can lie with data

  • Do all your maths

• For becoming manager

  • Start with small team

  • Technical understanding is important

  • Learn from sales-person for other part of business

  • Count of people doesn't matter

    • Developing skill is important

  • Keep 5-10% of time for learning per week

  • Understand a model development can take time

    • Sometime it can take 6-8 months

• Data analysis

  • Interact with the people to understand data

  • Do data cleaning. It is almost never the case when data cleaning is not needed

    • Missing data

    • Normalization

    • Biasness

    • Way too small data set

  • Use visualization tools for cleaning

  • Each data point represents a human being

    • See if the data is truely representation population. Or else it will be biased

  • Stakeholder handling

    • Folks may pressurize to get the insight they want

      • You need to decide if you like to take such analysis with less harm or back out

  • There are lot of ways

    • to view/interpret data

    • To clean data

• Decision making

  • Balance between release and experimentation

    • Data scientists sometime try to drill too much deep.

    • Need to balance with the release need

  • To present, seeing is believing

    • Use some graph etc

  • If data set is pretty small, work to enhance the size

  • measure of success

    • Giving value to the customer/company

•