Time-series Analysis with Python

1. Sources

The material of the course is based on the UCL lecture notes as taught by Prof. Alexandros Beskos and from various textbooks.

Sources:

1. Time Series Analysis: With Applications in R (SE); Cryer D. Jonathan, Chan Kung-Sik; 2009
2. Time Series: Theory and Methods (SE); Brockwell J. Peter, Davis A. Richard; 1991
3. Forecasting; Beskos Alexandros; 2016

2. Introduction

Time series is a sequence of consecutive and equally spaced observations measured over a period of time [3]. Forecasting is the academic field that aims to analyse the behaviour of these observations and infer statistical results that will help us predict future observations with the minimum error.

Mathematically, we can depict these observations as a stochastic process Y ={Y₁, ..., Y_i, ..., Y_N}, with 1:N being the discrete time, and then we can further break down the Y_i into trend, seasonal behaviour and white noise (random effect).