#sdsc6012 #course information

English / 中文

Course Overview

Course Code: SDSC6012
Course Name: Time Series and Recurrent Neural Networks
Academic Year/Semester: 2024/25 Semester A
Instructor: Prof. Linlin Wang
Email: (To be announced)
Office: (To be announced)

  • Lecture Time: (To be announced)

  • Consultation Hours: (To be announced)

  • Teaching Mode: Face-to-face

  • Tutor(s): (To be announced)

Abstract

In macroeconomics and other areas of business, science, and engineering, a significant amount of data is available as time series datasets. This course equips students with statistical tools to analyze such data, applying them to real-world problems using the R software. Students will review basic stochastic processes and time series concepts, then expand their knowledge to ARMA models, estimation methods, forecast properties, and GARCH models for volatility. The course also introduces recurrent neural networks for time series forecasting. A practical focus is maintained throughout, with hands-on data analysis using R.

Assessment Scheme

Component Weight Details
Test 25% Assesses conceptual description of statistical methods and RNNs for time series. Closed book.
Assignments 25% Assesses ability to use R for time series analysis, explanation, and presentation of results. GenAI Allowed.
Final Examination 50% 2-hour exam covering all intended learning outcomes, focusing on conceptual description and correct application of techniques. Closed book. Minimum passing requirement: 30% of the exam mark.

Intended Learning Outcomes (CILOs)

  1. Describe AR, MA, ARMA, ARCH, GARCH models, and recurrent neural networks for time series data. (20%)

  2. Apply time series models to analyze real data using R. (20%)

  3. Explain model selection criteria for time series models. (20%)

  4. Apply the models for time series forecasting using R. (20%)

  5. Apply recurrent neural networks to forecast time series data. (20%)

Learning and Teaching Activities (LTAs)

Type Brief Description CILOs Hours
Lectures Primary engagement through lectures, mini-lectures, and group exercises to consolidate conceptual understanding and application. 1, 2, 3, 4, 5 26 hrs/semester
Tutorial Exercises Team-based exercises to discuss and apply statistical tools through practical problem-solving. 2, 4, 5 13 hrs/semester

Keyword Syllabus

  • Autoregressive (AR), Moving Average (MA), Autoregressive Moving Average (ARMA) models

  • Parameter estimation

  • Model selection criteria

  • Properties of forecasts

  • Modelling volatility using ARCH and GARCH

  • Artificial neural networks

  • Recurrent neural networks (RNN)

  • Long Short-Term Memory (LSTM)

Reading List

Compulsory:

  1. Shumway, R. H., & Stoffer, D. S. (2017). Time Series Analysis and Its Applications: with R examples. Springer.

  2. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.

Additional:

  1. Brockwell, P. J., & Davis, R. A. (2016). Introduction to Time Series and Forecasting. Springer.

  2. Chollet, F., & Allaire, J. J. (2018). Deep Learning with R. Manning Publications.