Instructor: Mike Bowles, PhD
Location: Hacker Dojo (also available as webcast and recording)
Dates: Oct 12, 26, Nov 2, 9
Time: 9:00 am till 1:00 pm
Registration: Early Bird registraion 5 days before class starts $295 (for all 4 classes) through timeseriesmodels.eventbrite.com or by PayPal (mike at mbowles dot com). Regular registration $345 through eventbrite, PayPal or in class by check or cash.
This course will cover the basics of time series analysis and prediction. In particular, we'll address topics used in conjunction with financial time series and topics currently being adapted to handle large volume streaming data. The class will employ examples in R-code. The main text for the course will be Analysis of Financial Time Series by Ruey S. Tsay. Here's an outline.
1st Week Background for Time Series
Definitions, Stationarity, Tests, Examples
2nd Week Linear Models
General Linear Model, Fitting MA, ARMA, ARIMA etc. ARCH and GARCH volatility models
3rd Week Other Topics in Financial Time Series
Ito calculus and Black-Scholes model for option prices, Market microstructure, Co-integration and Pairs Trading, Basket trading
4th Week Kalman Filters and Non-parametric models
Simple Derivation of Kalman Filter, General Form of Multivariate KF, Identifying KF State Space Model from Time Series Data, Singular Spectrum Analysis and Change Point Detection