Business Analytics: Data-Driven Decisions

• Business analytics involves using data, statistical analysis, quantitative methods, explanatory and predictive models, and fact-based management to inform decisions and actions.
• Data-driven decision-making, statistical and quantitative analysis, predictive and explanatory modeling, and fact-based management are key elements of business analytics.

Business Analytics in Organizational Context

• Business analytics works with internal and external data sources, leveraging analytics to generate value.
• Descriptive analytics summarizes historical data.
• Predictive analytics uses models to forecast future outcomes.
• Prescriptive analytics recommends actions based on predictions.
• Big data assets are data-driven resources that fuel analytics.
• Actionable insights from data, big data technologies for storage and processing, data quality, and cloud computing are key components.

The Four V's of Big Data

• Volume refers to large datasets that require storage and analysis.
• Velocity refers to continuous streams of information requiring rapid processing.
• Variety means structured, semi-structured, and unstructured data from various sources.
• Veracity is the uncertainty and inconsistency in data, requiring techniques to improve realiability.

Building a Business Analytics Development Function

• A data scientist team, alignment with business goals, suitable tools/techniques, and a structured analytics methodology are key components.

The Role of a Data Scientist

• Data scientists extract value from messy, unstructured data using analytics and machine learning.
• Data scientists are described as a hybrid of data hacker, analyst, communicator, and trusted advisor.
• Programming, data wrangling, modeling, visualization, technology expertise, and communication are key skills.

Data Scientist Tasks

• 67% of data scientists perform basic exploratory data analysis.
• 61% of data scientists conduct data analysis to answer research questions.
• 58% of data scientists communicate findings to business decision-makers.
• 53% of data scientists perform data cleaning.
• 49% of data scientists develop prototype models.
• 47% of data scientists create visualizations.
• Other responsibilities include project organization, feature extraction and engineering, implementing machine learning algorithms, ETL processes, and developing dashboards.

Types of Business Analytics

• Descriptive analytics summarizes past events and what is currently happening.
• Diagnostic analytics answers "why" things happen through root cause analysis.
• Predictive analytics uses data to predict future trends and behaviors.
• Prescriptive analytics determines the best course of action among multiple options.

Modelling Techniques: Supervised vs. Unsupervised Learning

• The key difference between supervised and unsupervised learning involves the output variable being pre-defined/specified.
• Supervised learning is a machine learning approach where the model learns from labeled data (input-output pairs).
• Unsupervised learning is a machine learning approach where the model identifies patterns in data without labeled outputs.

Machine Learning Algorithms

• Classification assigns input data to predefined categories, like spam detection.
• Clustering groups similar data points without predefined labels, such as customer segmentation.
• Regression predicts continuous numerical values, like forecasting house prices.
• Anomaly detection identifies unusual data points, like detecting fraudulent transactions.

The Analytics Process

• The steps are: defining business objectives, collecting data, preparing and exploring data, creating training and test datasets, building and improving the model, and deploying the model.
• Data is split into training and test datasets to avoid overfitting the model.

Analytics Methodology

• Analytics methodology is a structured framework used to plan, develop, and control analytics solutions.
• A common framework is CRISP-DM (Cross-industry Standard Process for Data Mining).
• Phases of CRISP-DM include business understanding, data understanding, data preparation, modelling, evaluation, and deployment.

Business Objectives in Analytics

• Effective business objectives should be specific, measurable, relevant, achievable, and time-bound.
• Examples include improving response rates, increasing order sizes, and achieving high customer review scores.

Business Analytics Applications

• Key areas include Marketing, HR, Finance, and Procurement Analytics.

Finance Analytics

• Finance analytics is applied to fraud detection. For example, Machine learning identifies faked loan documents at NAB.
• It is applied to AI-based security. For example, CBA uses AI to detect and prevent scams.

Marketing Analytics

• Marketing analytics can be used for customer sentiment analysis such as Woolworths using LLMs to analyze customer feedback.
• It is used for personalization and efficiency such as Coles leveraging AI to enhance customer experience and improve store efficiency.

HR Analytics

• HR Analytics is applied to AI for talent acquisition by matching candidates to job roles.
• It is applied to AI-driven onboarding by enhancing the process to improve HR efficiency.