A textbook of Time Series Analysis

#1
I obtained a Bachelor Degree in Statistics in 1988. I was separated form this profession for more than 17 years because I was exercising Information Technology. Now I returned to Statistics and an revising the Statistics and Mathematics courses of my undergraduate study in order to revive my knowledge obtaind then. I have been searching in vain for a textbook of Time Series Analysis with mathematical rigour. I want to buy a book with proves and in which the association of formulas is shown by derivation. Is there anybody who knows such a book of Time Series Analysis and can recommend me? Thank you in advance for the trouble you take to help me.
 
#3
This book is for graduate students

I think Hamilton's book is what you're looking for

> http://press.princeton.edu/titles/5386.html
Thank you for your recommendation. But this book is specifically written for graduate level. It would be an appropriate book if I were doing a graduate study. I may need this book in the future. But for now, I am looking for a book at the undergraduate level which is written for a first Statistics course in Time Series Analysis and Forcasting.
 
#4
Tough treading really. For a first statistics course though the only shot you have really is in a linear regression book for undergraduates. I used "Hamilton, Regression with graphics", but I would probably use one aimed at R if I was learning on my own.

However, without experience with a simplier statistics topic and without a teacher its a tough bill. Its not that you can't learn it on your own its that it is essentially the 3rd topic in a series of proficiencies:
[0) probability might be inserted here]
1) general statistics
2) linear regression
3) time series

You have to start at the top of the chain and work down even learning on your own to be doing anything better than hacking at it.
 

mp83

TS Contributor
#5
To be honest as a graduate student I never of Hamilton's book as the first choice as I am in favor of more applied books,like using R series. You asked for derivations and proofs though,so there you go.Most books don't have proofs for serious things so Hamilton is oneway.

I personally heve read through its entity only Mills' Econometric Analysis of Financial Time Series. I wouldn't recommend it! Chatfield's and Time Series Analysis - With Applications in R.

Mostly because of his second book,I'll take a chance to Kirchgaessner's Introduction to Modern Time Series Analysis.

For Regression there;s no reason not to get the FREE book
> http://cran.r-project.org/doc/contrib/Faraway-PRA.pdf
 
#7
Oh I totally second course notes. As an undergraduate I learned so much from peoples notes online. I have funneled that energy into papers as a graduate student, but I am all for the practice. You can quickly find slides with a half dozen variations on the topic etc. For most common things more examples than you can count etc. There is a UCLA website for statistics too that is handy.

Offtopic.

Hehe I can't say that I honestly recommend Hamilton either. I think my instructors course notes helped more at that point than Hamilton. Its just the only one I have because it is not really my area. My officemate did wavelets this semester and I was like "no thanks....". Though I just got Elements of Statistical Learning by Hastie (not an intro text and not really about time series), and I been reading through it this week to get a sense of whats there, basically like a book instead of a textbook. Anyway he passes kernal methods and I thought "these are kind of cool". So who knows...

Mostly though I like statistical design topics. I am so jealous of these graduate students in agriculture, pyschology, etc that are so good at certain applied designs. That is my goal: to match their knowledge of common designs. I almost wish I hadn't got a contract to do research this summer so I could drive at it more directly.

The instructor here with the most experience in applied GLM designs says a stat student interested in those things should focus on breadth of experience rather than intense expertise in one design topic.
 
#8
To be honest as a graduate student I never of Hamilton's book as the first choice as I am in favor of more applied books,like using R series. You asked for derivations and proofs though,so there you go.Most books don't have proofs for serious things so Hamilton is oneway.

I personally heve read through its entity only Mills' Econometric Analysis of Financial Time Series. I wouldn't recommend it! Chatfield's and Time Series Analysis - With Applications in R.

Mostly because of his second book,I'll take a chance to Kirchgaessner's Introduction to Modern Time Series Analysis.

For Regression there;s no reason not to get the FREE book
> http://cran.r-project.org/doc/contrib/Faraway-PRA.pdf
Thanks a lot though a book on regression I ordered is already on its way to me. It is not bad to have more than one sourse.
 
#9
Tough treading really. For a first statistics course though the only shot you have really is in a linear regression book for undergraduates. I used "Hamilton, Regression with graphics", but I would probably use one aimed at R if I was learning on my own.

However, without experience with a simplier statistics topic and without a teacher its a tough bill. Its not that you can't learn it on your own its that it is essentially the 3rd topic in a series of proficiencies:
[0) probability might be inserted here]
1) general statistics
2) linear regression
3) time series

You have to start at the top of the chain and work down even learning on your own to be doing anything better than hacking at it.
Well, I am not new to statistics and mathematics and statistics and mathematics are not new to me at all. I graduated in statistics in 1988. The only reason why I am looking for textbooks covering the curriculum of my undergraduate study is to revive my knowledge of statistics and mathematics which I partially lost because I was exercising Information Technology for the last 18 years. Therefore, I don't need any instructor for courses I already attended in class during my undergraduate study. I already began revising both statistics and mathematics side by side starting from the elementary as you said. I am doing very well upto now and will certainly achieve my goal within the time limit I set. Thank you for your advice.