module specification

This module begins with a foundational level of mathematics and python appropriate for non-cognate entry to a masters in Economics and Data Analytics. The mathematics will be primarily of an applied nature including linear algebra (matrices). The python will begin with use of Jupyter Notebook, handling of data and simple functions. Subsequent to this, we will use python to acquire data via to public API, manipulation of data and the running of a predictive model, for example Long Short Term Memory (LSTM). At the same time as establishing python skills, the module will provide students with a solid grounding in the application of data for business and the use of analytics to create commercial value.

The aims of the module are:

- You will have introductory skills in Python programming language.

- You will be fluent in the differences between types of data and how to handle them.

- You will be able to translate real world challenges into data and models, and vice versa.

- You will be able to construct at least one type of predictive model.

- You will be able to put forward data driven results, solutions and proposals.

Introduction to mathematics and python. A foundational level of maths and statistics appropriate for non-cognate entry to economics and data analysis. Math topics will include algebra and matrices. Introduction to Python and how to access and manipulate data. Use of Jupyter Notebook (LO1).

Commercial applications of data. An overview of data types and purposes with specific reference to economic, social and business problems. This will involve the analysis of appropriate case studies. Seminar activities will require students to use new data structures or perform new actions to data. Connections between the real world and quantitative data will be a core focus (LO2, LO3).

Use of either Python or an econometric software package (e.g., R or Eviews) for predictive modelling. In Python this may be Long Short Term Memory (LSTM) and in R this may be the use of Principal Component Analysis (PCA) within a multivariate regression (LO4).

Data analysis to solve problems. Students will apply their findings from the data analysis and predictive modelling to the original real world problem. Based on this, participants will lay out potential insights, proposals and solutions (LO5)

Learning consists of ‘formal’ classroom learning directed by the teaching team, and reflective independent learning. The formal learning involves 12 weeks of 1 hour of lectures, and 2 hour of seminars where computer-based exercises will be attempted. Independent learning includes reading of the course material, and working on weekly exercises. Formal teaching will focus on application over theory.

Students will receive presentation slides and weekly exercises via Weblearn. Exercises will be used to develop student knowledge and where appropriate require use of Jupyter Notebook (Python) and a statistical software package chosen by the module leader (e.g., Eviews or R). A portion of the weekly exercises will be reviewed in class. Formal teaching is to be accompanied by 8 hours a week of independent study. 68 hours are allocated to the completion of a summative coursework assessment.

Students will be grouped into action learning sets of 4-6 persons and will be given time during seminars to problem solve together. They will be encouraged to maintain a journal of experiences in order to reflect on their personal development.

On the completion of this module students will be able to:

1. Use python programming language to access and manipulate data.

2. Understand of the differences between time series, cross sectional and panel data, and how these differences influence economic and predictive modelling.

3. Relate data, models and predictions to real world, commercial and academic problems.

4. Construct one predictive model, either in Python or an econometric software package (e.g., R or Eviews).

5. Conclude your analysis by proposing data driven solutions or recommendations.

http://EC7096 Modelling of data and predictive analytics (subject to validation) | London Metropolitan University (talis.com)