TODO6Jun2020

TODOJUN2020 - TS VERSION
STL IMPLEMENTATION USING:
REF 78/ - UNDERSTAND, THEN IMPLEMENT
REF 79/ - READ AND UNDERSTAND THEN IMPLEMENT
REF 80 - DONE

78-REFS
PCA + ALPHA HULL
R. J. HYNDMAN. COMPUTING AND GRAPHING HIGHEST DENSITY REGIONS. AMER.
STATIST., 50(2):120–126, 1996
PCA + HDR
B. PATEIRO-LOPEZ AND A. RODRıGUEZ-CASAL. GENERALIZING THE CONVEX HULL ´
OF A SAMPLE: THE R PACKAGE ALPHAHULL. J. STAT. SOFT., 34(5):1–28, 4 2010

[79] REFS -
FOR STL-
1. FOURIER TRANSFORM [2] -
 P. BLOOMFIELD. FOURIER ANALYSIS OF TIME SERIES: AN
INTRODUCTION. JOHN WILEY & SONS, 2004
 OR WAVELET TRANSFORM [19] - 
 D. B. PERCIVAL AND A. T. WALDEN. WAVELET METHODS FOR
TIME SERIES ANALYSIS, VOLUME 4. CAMBRIDGE UNIVERSITY
PRESS, 2006 
WHEREAS, FOR NON-PARAMETRIC METHODS, THE BASIS IS DATADRIVEN [18] -
 V. MOSKVINA AND A. ZHIGLJAVSKY. AN ALGORITHM BASED
ON SINGULAR SPECTRUM ANALYSIS FOR CHANGE-POINT
DETECTION. COMMUNICATIONS IN STATISTICS-SIMULATION
AND COMPUTATION, 32(2):319{352, 2003

ARIMA [26] - W. W.-S. WEI. TIME SERIES ANALYSIS. ADDISON-WESLEY PUBL, 1994.
EXPONENTIAL SMOOTHING[11] - R. H. JONES. EXPONENTIAL SMOOTHING FOR MULTIVARIATE TIME SERIES. JOURNAL OF THE ROYAL STATISTICAL SOCIETY. SERIES B (METHODOLOGICAL), PAGES 241{251, 1966
KALMAN FILTER [9] - S. S. HAYKIN, S. S. HAYKIN, AND S. S. HAYKIN. KALMAN FILTERING AND NEURAL NETWORKS. WILEY ONLINE LIBRARY, 2001.
STATE SPACE MODELS [6] - J. DURBIN AND S. J. KOOPMAN. TIME SERIES ANALYSIS BY STATE SPACE METHODS. NUMBER 38. OXFORD UNIVERSITY PRESS, 2012


INSTALL R STUDIO
COLLECT TSFEATURES FOR ALL YAHOO DATASETS.
IMPLEMENT PCA ON THE FEATURES.
TRY ANOMALOUS PACKAGE IMPLEMENTATIONS OF THE ALPHA HULL AND THE DENSITY BASED OUTLIER DETECTION ALGORITHMS ON THE PCA OUTPUT DATA.



LIST HAJARS REFERENCES - READ AND UNDERSTAND - IMPLEMENT
AR
MA
ARIMA
[67] R. Adhikari and R. K. Agrawal, An Introductory Study on Time Series Modeling and Forecasting. LAP LAMBERT Academic Publishing. AR
[68] “Autoregression Models for Time Series Forecasting with Python,” https://machinelearningmastery.com/ Accessed (23/10/2019). AR
[69] G. A. F. Seber and A. J. Lee, Linear Regression Analysis, 2nd ed. Wiley. AR
[70] I. J. Myung, “Tutorial on maximum likelihood estimation,” vol. 47, no. 1, pp. 90–100.AR

--------------
18/6/20
Reinstalled R and R studio
Learning data manipulation with R using the tutorial at -https://paldhous.github.io/ucb/2016/dataviz/week7.html
installed oddstream
installed tsfeatures
looking for wavlet/fourier transform time series

1. STL decomposition
	a. STL for outlier detection - threshold calculation value
2. Time series modelling 
	a. AR
	b. MA
	c. ARIMA
	d. SVM
	e. MLPs
	f. SANN
	g. LSTM

model a ts to a func f, then predict value Rt as R't using f and calculaitng pred error pe = Rt - R't
outlier if pe > fixed threshold value
f has parameters - estimated using stochastic methods or ml
stochastic methods use probability models 
AR - ts = [Rt] such that Rt is the sum of prev data records plus an error
The relationship between data records is called correlation

machine learning mastery ar tutorial