DALMINE
Overview
Date/time interval
Syllabus
Course Objectives
The course aims to bridge the gap between technical data analysis and strategic decision-making in a business environment.
At the end of the course students will be able to:
1. Prepare the data: Manage the acquisition of data from different sources and their cleaning to make them reliable.
2. Modeling reality: Apply predictive statistical methodologies (Regression and Classification) to anticipate future scenarios rather than just describing the past.
3. Support decisions: Transform mathematical outputs into actionable business insights for Marketing, Operations, and Strategy functions.
4. Create decision-making tools: Develop interactive interfaces (Data Apps) that allow stakeholders to simulate scenarios in real time.
Course Prerequisites
Basic knowledge of programming (Python), fundamentals of SQL (simple queries) and descriptive statistics, acquired during the bachelor degree course. The fundamental concepts will be reviewed in a review lesson.
Teaching Methods
Teacher’s lecture slides and Python-based coding exercises
Assessment Methods
The exam consists of two parts:
Written test (compulsory):
Evaluation: up to 30/30 (in the absence of a project).
Optional Project (Bonus/Partial Exemption):
Carrying out a group project on a real dataset. The project involves the implementation of a complete pipeline (Cleaning → Interactive App → Template).
The project guarantees a score that is added to the result of the theoretical part.
Evaluation based on: methodological correctness and "Data Storytelling" skills
Contents
MODULO 1: Data Engineering & Data Quality
1. Heterogeneous Source Management: The challenges of integrating static data (e.g., manual Excel/CSV reports) with dynamic data (e.g., extracts from SQL Databases). The concept of "Single Source of Truth".
2. Data Quality Management: Impact of data quality on decisions (GIGO: Garbage In, Garbage Out). Strategies for dealing with missing data (Statistical Imputation vs Elimination) and input errors.
3. Feature Engineering: Transform business information into numerical variables for algorithms. Encoding of categorical variables, date management, normalization (Scaling) and discretization.
MODULO 2: Predictive Analytics & Modeling
1. Regression Analysis: Linear (Multiple) Regression Managerial interpretation of coefficients as decision-making levers ("If I increase X, how much does Y change?"). Verification of assumptions and significance (P-value, R-framework).
2. Forecasting (Time Series): Analysis of the components of demand (Trend, Seasonality). Forecasting techniques applied to demand planning.
3. Classification: Model binary events (Yes/No, Churn/Loyal). Logistic Regression and Decision Trees.
4. Model Evaluation: Beyond Accuracy. Confusion Matrix, Precision, Recall. Economic analysis of False Positives vs False Negatives.
MODULO 3: Applied Business Analytics
1. Marketing Analytics: Customer segmentation through Clustering (K-Means) and RFM (Recency, Frequency, Monetary) behavioral analysis.
2. Operations & Supply Chain: Using Forecast Error (RMSE) for Statistical Calculation of Safety Stock and Analysis of Inventory Priorities (ABC Analysis).
3. HR Analytics (outline): Predictive analysis of staff turnover.
4. Unstructured Data & NLP: Text analytics for business (reviews, complaints). Practical Sentiment Analysis (lexicon-based). Managerial introduction to the use of Generative AI APIs (LLMs) in business processes: opportunities and costs. (very introductory)
MODULO 4: Data Products & Interactive Reporting
a. Design of Decision Support Systems: Principles of Effective Visualization.
b. Interactive Computing: Evolution from static report to Web App. Introduction to Python libraries for creating interfaces (Streamlit).
c. What-If Analysis: How to connect a predictive model to a user interface to simulate alternative scenarios.