TOKT11122 — Time Series Analysis and Forecasting

Số tín chỉ: 3
Giờ tự học: 90
Khoá hiển thị: K68-2026

Mô tả học phần

This course provides an introduction to time series analysis, highlighting its applications in the stock market and includes setting up Anaconda Python and Jupyter. Starting with NumPy, students will learn about arrays, indexing, and operations, applying these to create stock market indicators such as SMA, EMA, MACD, and RSI through hands-on exercises. The focus then shifts to Pandas for data handling and analysis, teaching students about Series, DataFrames, and data visualization, with practical exercises on stock market data. Students will explore visualizing financial time series and delve into advanced analysis with Statsmodels, covering error, trend, and seasonality decomposition. The curriculum includes forecasting models, from basics to complex concepts like ARIMA, SARIMA, and GARCH models, complemented by code-along sessions. By the course's end, students understand how to apply time series in analyzing and forecasting stock market trends with Python's powerful libraries, ready to apply these skills in financial contexts.

Yêu cầu của học phần

Ensure at least 80% attendance in the class.

Achieve a participant score of 5.0 or higher.

Must join projects

Actively contribute to the lesson.

Engage with the textbook and complete projects.

Avoid eating or using cell phones during class.

Use of laptops and iPads is permitted only for practice and calculation purposes.

Kế hoạch giảng dạy

Tuần 1

Lecture 1. Course Introduction Introduction to Time Series Analysis and Forecasting course 1.2. Group project requirements for the course. 1.3. Setup working environment by installing Anaconda Python and Jupyter 1.4. Time Series Concepts and Python

Đánh giá

Discuss

Tuần 2

Lecture 2. NumPy, Pandas and Its Applications for Time Series Analysis 2.1. Review NumPy 2.2. Review Pandas 2.3. Explain and implement indicators in stock market by Python 2.4. Exercises

Đánh giá

Discuss

Tuần 4

Lecture 3. Exploratory Time Series Data Analysis 3.1. Partial Autocorrelation Functions 3.2. White Noise Test 3.3. Simple Time Series Compositions 3.4. Time Series Decomposition and Smoothing

Đánh giá

Discuss

Tuần 6

Lecture 4. Stationary Time Series Models 4.1. Backshift Operator, Differencing, and Stationarity Test 4.2. Moving Average Models 4.3. Autoregressive Models 4.4. Autoregressive Moving Average Models 4.5. Time Series Analysis with Statsmodels 4.6. Exercises and Implementations

Đánh giá

Discuss

Tuần 8

Lecture 5. ARMA and ARIMA Modeling and Forecasting 5.1 Model Building Problems 5.2. Estimation Methods 5.3. Order Determination 5.4. Diagnosis of Models 5.5. Forecasting 5.6. Exercises and Implementations

Đánh giá

Discuss

Tuần 10

Midterm Exam

Đánh giá

Homework and Midterm 1

Tuần 11

Lecture 6. Nonstationary Time Series Models 6.1. The Box-Jenkins Method 6.2. SARIMA Model Building 6.3. REGARMA Models

Đánh giá

Discuss

Tuần 13

Lecture 7. Financial Time Series and Related Models 7.1. Stylized Facts of Financial Time Series 7.2. GARCH Models 7.3. Other Extensions

Đánh giá

Discuss

Tuần 15

Group project report

Đánh giá

Assessment sheet with the following criterium: (i) Content; (ii); (iii) Time; (iv) Cross-evaluation

Học liệu

Introduction to Modern Time Series Analysis — Gebhard Kirchgässner, Jürgen Wolters, Springer, 2013
Python for Data Analysis — Wes McKinney, O’Reilly Media, 2022
Python for Finance Cookbook — Eryk Lewinson, Packt Publishing, 2022
Time Series Analysis with Python Cookbook: Practical recipes for exploratory data analysis — Tarek A. Atwan, data preparation, 2022
Applied Time Series Analysis and Forecasting with Python — Changquan Huang and Alla Petukhina, Springer, 2022
Machine Learning for Time-Series with Python — Ben Auffarth, Packt Publishing, 2021
Machine Learning for Time Series Forecasting with Python — Francesca Lazzeri, Wiley, 2020
https://online.stat.psu.edu/stat510/