Business Analytics Methodologies Quiz

Business Analytics Methodologies

• Descriptive: Analyzing past and present data to understand trends and patterns.
  • Example: A pie chart breaking down a company's customer demographics.
• Diagnostic: Helps identify the root cause of an event, providing insights into why a trend occurred.
  • Example: Analyzing a failed component on an assembly line to determine the reason for failure.
• Predictive: Uses existing data to identify patterns and predict future outcomes.
  • Example: Forecasting coat sales changes based on predicted winter temperature variations.
• Prescriptive: Uses data analysis to suggest the best course of action for a specific situation.
  • Example: Analyzing customer data to suggest effective marketing strategies.

Types of Metrics

• Discrete metrics: Values that can be counted using whole numbers, representing distinct and separate entities.
  • Examples: Number of customers, products sold, customer complaints, website visits.
• Continuous metrics: Values that can take on any value within a specific range and are measured rather than counted.
  • Examples: Revenue, profit margin, time spent on a website, average order value.

Scope of Data Mining

• Business & Marketing:
  • Customer segmentation: Categorizing customers based on shared characteristics.
  • Anomaly detection: Identifying data points that deviate significantly from normal patterns, such as detecting fraudulent credit card transactions.
  • Association rule learning: Discovering relationships between variables in a dataset, such as finding frequently purchased product combinations.
  • Clustering: Grouping similar data points together, like segmenting customers based on preferences or behaviors.
  • Classification: Predicting categorical labels for data points, such as determining whether a customer will stop using a product based on demographics and purchase history.
  • Regression: Predicting numerical values based on input variables, such as predicting house prices based on factors like size, bedrooms, and location.
  • Summarization: Condensing large datasets into a concise form, like generating a customer's purchase history summary.
• Government:
  • Public safety: Predicting crime hotspots and optimizing resource allocation.
  • Fraud detection: Identifying fraudulent activities in government programs.
  • Policy analysis: Assessing the impact of government policies on various sectors.
• Science & Research:
  • Scientific Discovery: Uncovering new patterns and insights in scientific data.
  • Drug Discovery: Identifying potential drug candidates and understanding their interactions.
  • Climate Modeling: Analyzing climate data to predict future trends.

Data Mining Tasks

• Data Mining: The process of discovering patterns within data to uncover valuable information.

Data Mining Architecture

• Data Collection: Gathering data from various sources and storing it in a database or data warehouse.
• Data Mining: Processing the collected data to identify patterns using data mining engines.
• Pattern Evaluation: Analyzing and assessing the discovered patterns for their significance.
• Knowledge Extraction: Extracting valuable insights and knowledge from the patterns and storing them in a knowledge base.
• User Interface: Providing tools for visualizing and interpreting the results.

KDD Architecture

• Data Cleaning: Removing noise and inconsistent data from the dataset.
• Data Integration: Combining data from multiple sources.
• Data Selection: Retrieving relevant data for analysis from the database.
• Data Transformation: Transforming or consolidating data into forms suitable for mining by applying summary or aggregation operations.
• Data Mining: Applying intelligent methods to extract data patterns, such as using algorithms like Apriori, FP-growth, and GSP.