Technical Training

Data Analytics & Machine Learning

Objectives

The aim of the module is to enable the learner to program statistical, and in particular, machine learning applications, based on manufacturing data sets, using standard mathematical tools.

Programme overview
This module guides the learner in an examination of the application of statistics and experimental design to use cases in industry. The aim of the module is to enable the learner to program statistical, and in particular, machine learning applications, based on a variety of data sets, using standard mathematical and statistical tools.  

Programme description
The module will review the application of statistics and experimental design to applications in industry. Learners will be guided through the fundamentals of the R programming language and use analytical methods to identify, analyse and solve industry related problems. 

Course Outline  

• Runs every Wednesday for 12 weeks from 29 May to 21 August 2024
• Delivered fully online
• 90 minutes self directed learning each week
• Weekly tutorial

  Content includes

  On completion of this module the learner will/should be able to;

1. Analyse the application of computational intelligence to decision-making problems in a relevant industrial context.

2. Design an experiment to generate suitable datasets and apply statistical inference to extract valuable information.

3. Program a statistical software tool to perform time series analysis and forecasting.

4. Investigate approaches to machine learning, select, and develop appropriate algorithms for a specific data stream.

5. Implement a data analytics application for decision support on an empirical data stream situated in a manufacturing context, relevant to the learner’s industrial practice.

Operational Details 
The module starts with an introduction to the R language and RStudio, the environment used in the module. While the learners are taught to use R they also given a revision on the relevant statistical concepts that will be used throughout the course. Once they are familiar with the tool chain and the statistical background demanded by the course, the module moves on to covering linear methods for regression and classification problems, including data resampling techniques and linear model selection and regularisation. The module then shows how to change linear models so they can also accommodate non-linear relationships, as well as non-linear methods such as three based models and support vector machines. Include in the coverage of non-linear methods are unsupervised techniques such as features engineering and clustering. The last theoretical topic covered is artificial neural networks, and their development up to today’s deep learning applications is introduced. Finally, the module integrates the knowledge from the previous topics to develop a fully functional data analysis dashboard using the Shiny library, as well as demonstrating how to deploy it to a local or cloud server.        

Who should attend

Those seeking to implement digitisation in their organisation

Qualification

5 ECTS of a Level 9

Upcoming dates: