Time series analysis is a fascinating but extensive and potentially confusing topic in statistics. Typically, the idea is to use previous observations of a random variable to attempt to determine the nature of the stochastic process generating the observations and hence, predict future values of the variable. I will introduce basic concepts of the time-domain and the frequency-domain methods for analyzing time series, describe simplified exponential smoothing and forecasting, and touch on the more complex forecasting methods based on the so-called “state-space” representation. I will also draw attention to R, the high powered, zero cost software package that has essentially made statistics available to all.
About Joseph Ofungwu, Ph.D., P.E.
Dr. Joseph Ofungwu is an environmental engineer with two decades of multi-faceted experience in environmental consulting and related water resources fields. His educational background is diverse, with a BS in civil engineering, MS in hydraulics and water resources, and Ph.D. in civil/environmental engineering. And, so is his professional environmental career, starting off with remedial investigations, remedial design and later expanding to chemical exposure risk assessment and pollutant transport analysis and management Through the course of his career, he has found statistics to be an indispensable tool for effective environmental practice, and recently published a “Statistics in Practice” book titled Statistical Applications for Environmental Analysis and Risk Assessment (Wiley, 2014) to share his experiences. He is currently exploring teaching opportunities and intends to write more on the subjects of risk assessment and pollutant transport.