UE Data analytics for industrial engineering - WGUS2092

Goal(s)

Part I
Students will learn about data preparation methods for machine learning, and how to integrate them into data science projects.
Students will know how to manage their data, sort it, and organize it efficiently. They will be able to present relevant visualizations of their data and results. They will have acquired the behavior of a responsible and ethical data engineer.

Part II
The course introduces the first tools for processing quantitative and qualitative data using machine learning.
The analysis methods that will be learned allow for automatic classification; construction of predictive models; evaluation of the performance of the methods; diagnosis of the limits of the applications of these methods.

Content(s)

Part I
.0 Introduction: Data Science Project Management
Steering data science projects, based on CRISP-DM
.1 Data handling & Data Engineer responsibilities (ethics, security, etc.)
.1.1 Technical data management
Data format, variable formats; basic operations (reads, writes; sorts; selections, projections, filters; merges)
.1.2 Technical management of results (visualization)
Types of graphics, principles of good visualization
Make and discuss technical choices and representations
.1.3 Societal management
Legal aspects (RGPD), sustainable aspects (risks on people [customer and staff] as well as environmental costs), security (who holds the data, spying...).
.1.4 Implementation : Micro-project

Part II
.1 Issues in machine learning, supervised machine learning (regression, classification)
supervised vs unsupervised methods (quick presentation of some unsupervised methods (k-means, dendrograms)).
.2 Regression and classification methods: linear regression and logistic regression; models, algorithms and resolutions
.3 Internal evaluation of regression and classification: Errors, residuals and prediction evaluation
.4 External Evaluation: Statistical Assumptions and Model Evaluation
.5 Implementation on different databases

Students will have taken and validated the following courses: Probability and Statistics; Programming with R and Python