Data Science Methodology
An Analytic Approach to Capstone
Project
What is Data Science
Methodology?
• A structured, iterative process used by data
scientists to solve problems.
• It consists of multiple stages from
understanding the problem to deploying a
solution.
Key Modules
• 1. From Problem to Approach – Define
problem, use 5W1H & Design Thinking.
• 2. From Requirements to Collection – Identify
& gather structured/unstructured data.
• 3. From Understanding to Preparation –
Analyze, clean & transform data.
• 4. From Modeling to Evaluation – Develop &
assess models.
• 5. From Deployment to Feedback –
Implement, refine & optimize.
Analytic Techniques
• - Descriptive Analytics: Identify trends &
patterns.
• - Diagnostic Analytics: Determine causes of
events.
• - Predictive Analytics: Forecast future
outcomes.
• - Prescriptive Analytics: Recommend best
actions.
Data Processing Steps
• - Collection: Primary (surveys, experiments) &
Secondary (books, web data).
• - Understanding: Assess quality using statistics
& visualization.
• - Preparation: Cleaning, integration & feature
engineering.
• - Modeling & Evaluation: Test models using
accuracy, precision, recall.
• - Deployment & Feedback: Implement &
refine based on results.

Condensed_Data_Science_Methodology-AI.pptx

  • 1.
    Data Science Methodology AnAnalytic Approach to Capstone Project
  • 2.
    What is DataScience Methodology? • A structured, iterative process used by data scientists to solve problems. • It consists of multiple stages from understanding the problem to deploying a solution.
  • 3.
    Key Modules • 1.From Problem to Approach – Define problem, use 5W1H & Design Thinking. • 2. From Requirements to Collection – Identify & gather structured/unstructured data. • 3. From Understanding to Preparation – Analyze, clean & transform data. • 4. From Modeling to Evaluation – Develop & assess models. • 5. From Deployment to Feedback – Implement, refine & optimize.
  • 4.
    Analytic Techniques • -Descriptive Analytics: Identify trends & patterns. • - Diagnostic Analytics: Determine causes of events. • - Predictive Analytics: Forecast future outcomes. • - Prescriptive Analytics: Recommend best actions.
  • 5.
    Data Processing Steps •- Collection: Primary (surveys, experiments) & Secondary (books, web data). • - Understanding: Assess quality using statistics & visualization. • - Preparation: Cleaning, integration & feature engineering. • - Modeling & Evaluation: Test models using accuracy, precision, recall. • - Deployment & Feedback: Implement & refine based on results.