Organizations and business are overwhelmed by the flood of data continuously collected into their data warehouses and arriving from external sources – the Web above all. Traditional exploratory techniques may fail to make sense of the data, due to its inherent complexity and size. Data mining and knowledge discovery techniques emerged as an alternative approach, aimed at revealing patterns, rules and models hidden in the data, and at supporting the analytical user to develop descriptive and predictive models for a number of business problems. This short course focusses on the main applications scenarios of data mining to challenging problems in the broad CRM domain - Customer Relationship Management.
|01.||Gio 11.04.2019 - 09:00-13:00||Introduction to data mining and big data analytics||slides: intro slides: case studies||Giannotti|
|02.||Gio 11.04.2019 - 14:00-18:00||Data understanding; data preparation; Knime tutorial||slides slides data understanding Tutorial Knime 01_titanic_data_understanding||Pedreschi, Guidotti|
|03.||Ven 12.04.2019 - 09:00-13:00||Clustering analysis & customer segmentation||slides clustering slides customer segmentation||Pedreschi|
|04.||Ven 12.04.2018 - 14:00-18:00||Clustering analysis: exercises with Knime||02_titanic_clustering||Pedreschi, Guidotti|
|05.||Lun 15.04.2019 - 09:00-13:00||Classification & prediction||slides classification Visual Introduction to Classification with Decision Trees||Pedreschi|
|06.||Lun 15.04.2019 - 14:00-18:00||Classification & prediction: exercises with Knime||05_titanic_classification||Pedreschi, Rossetti|
|07.||Mar 16.04.2019 - 09:00-13:00||More on Classification: from decision trees to deep learning||Evaluation of classifiers KNN & Naive Bayes Neural Networks & SVM Ensemble methods & Wisdom of the crowd||Pedreschi|
|08.||Mar 16.04.2019 - 14:00-18:00||Classification & prediction: exercises with Knime. Project work||Giannotti, Rossetti|
|09.||Mer 17.04.2019 - 09:00-13:00||Pattern and association rule mining & market basket analysis||5.dm-ml_patternmining-2018.pdf||Giannotti|
|10.||Mer 17.04.2019 - 14:00-18:00||Pattern and association rule mining: exercises with Knime||03_titanic_pattern 04_coop_pattern||Giannotti, Rossetti|
|11.||Gio 18.04.2019 - 09:00-13:00||Case studies. Prediction models for promotion performance and churn analysis||5.dml-ml-exemplarproject-churn-fraude-.pdf 5.dm_ml_exemplarprojects-shoppingbehaviour_innovators.pdf||Giannotti|
|12.||Gio 18.04.2019 - 14:00-18:00||Hints on data science with Python. Data Science Privacy & Ethics.||5.dml-ml-privacy_etica-.pdf||Giannotti, Rossetti|
3. Telco Churn. (for details see http://didawiki.di.unipi.it/doku.php/dm/mains.santanna.dm4crm.2016)
5. Credit Card (for details see https://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients)
Each group (2-3 people) is required to deliver a report (max 20 pages including all figures) describing the methods adopted and the discussion of the most interesting achieved results with reference to the tasks listed below. Assume that the report is targeted to a marketing strategist, who is interested to learn the story inferred in the various data mining analyses and to receive suggestions on how to take appropriate actions as a consequence.
1. Data Understanding: useful as a preliminary step to capture basic data property. Distribution analysis, statistical exploration, correlation analysis, suitable transformation of variables and elimination of redundant variables, management of missing values.
2. Pattern Mining Analysis. Problem: prepare data and extract interesting association rules and frequent patterns. The report should discuss the parameters used for the analyses, justifying your findings related to the most interesting rules according to the different measure introduced in the course.
3. Customer Segmentation. Problem: find a high-quality clustering using clustering algorithms and discuss the profile of each found cluster (in terms of the properties that describe the properties of the customers of each cluster). The report should illustrate the adopted clustering methodology and the cluster interpretation. In particular, in case of k-means, it is necessary to discuss the identification of the best value of k and the characterisation of the obtained clusters by using both analysis of the k centroids and comparison of the statistics of variables within the clusters with that in the whole dataset.
4. Classification Analysis. Problem: find a high-quality decision tree for predicting a feature of a customer. The report should illustrate the adopted classification methodology and the decision tree validation and interpretation, describing also the process adopted to select the proposed tree, together with its quality evaluation.
Deadline: send the report by email to all instructors within 22 June 2019. Specify [MAINS] in the subject of the email.
The exam consists in the evaluation of the report of the proposed mining exercises.