A course on understanding and implementing time series methods.
All processes ride on time. Therefore, time series appears in all fields -- science and engineering, business and economics, social sciences. Understanding how to handle this dimension is interesting and valuable.
This is a course on understanding the methods, and how to implement them correctly. For better illustration of concepts, we will rely on examples from the domain of financial markets. Basic statistics is necessary. Knowledge of finance is not necessary.
We will understand why time series is different from traditional statistics (based on IID assumptions). We will cover traditional time series models and also new models based on machine learning algorithms. We will cover some advanced concepts as well -- like cointegration, and volatility modeling. Time series observations need not be in regular intervals -- so, we will see some traditional approaches to analyse such processes.
You will encounter time series on several occasions. Having done this course, you will be able to use the methods in different areas or domains. You will understand the assumptions that underlie different models, how to choose appropriate strategy, and also how to implement the models. Time series can address causality in specific situations; it is extensively used for predictive modelling. All concepts will be relevant and helpful for both research in time series and applications.
Lectures will be available in due course.