Questions tagged [time-series]
Time series are data observed over time (either in continuous time or at discrete time periods).
11,564
questions
1
vote
0answers
13 views
Bayesian Model Averaging (Bayesian Averaging of Classical Estimates) Issue
I am trying to implement a method used by this paper (Described briefly at the bottom of page 3):
From what I understand, it does 2$^k$ OLS regressions each time step to forecast the next time step ...
0
votes
1answer
18 views
Cointegration: Long vs Short Run
I am running an ECM model that has one cointegrating vector but two stochastic trends within the cointegration vector ex.(1, -1, 1). Can I use the cointegration vector inside an error correction model ...
0
votes
0answers
29 views
How to approximate a normal distribution?
I want to reduce the space requirement.
I have a main factory with a psychic alarm/monitor that can predict accidents.
It pings once per second with continuous variables of peak volume, pitch, and ...
0
votes
0answers
6 views
Numeric Differentiation of timeseries function f(x,t) affected by fixed time series input
I am trying to calculate the Hessian in order to preform parameter estimation using Hamiltonian/Quasi Newton MCMC. Before using this on the complicated model I wanted to test it on something easy. As ...
0
votes
0answers
12 views
Multiple Time Series Anomaly Detection
I have several different unlabeled multivariate time series, each drawn from what I believe to be the same distribution, but virtually independent from each other. Each has a different length (in time)...
0
votes
0answers
11 views
Is there a statistics method similar to Anova for groups of data that change over time (or other independent variable)?
My dependent variable (temperature), is time dependent (time groups 1-8, each 15 minutes apart). 5 different plants were measured. I want to know if there is a significant difference between the ...
0
votes
0answers
13 views
Two conceptions of including linear trend
Let's take panel data for which I want to include linear trend :
...
3
votes
0answers
28 views
Burg's method for estimation of AR-models
I am trying to get an intuitive understanding of Burg's method for estimation the coefficients of an AR-model. Say we have an AR(1)-process with
\begin{equation*}
X_t = a X_{t-1} + \varepsilon_t
\...
0
votes
0answers
6 views
Extracting average from multiple time series that starts at different times?
I have many measurements where I would like to extract the Average of the "flat" area of the below graph.
My challenge, is that the starting of the flat section of my data is different for ...
0
votes
0answers
15 views
the expected value of Autoregressive model with linear trend
In the stationary time series like the following with intercept and AR(p) model, the expected value is the
intercept / (1-AR(1)-AR(2)) = mean
##B0 = alpha(intercept)
Then, I include a trend to data ...
0
votes
0answers
12 views
Discriminating model and data forecasting error
My process looks like this
time series 1 --> model 1 --> time series 2 --> model 2 --> time series 3
An initial time series, which is a forecast, is ...
1
vote
0answers
12 views
ARCH(2) Process Stationarity
Lets say that we have an ARCH(2) process such that:
$$ y_t = \epsilon_t \sigma_t$$ where $\epsilon_t \textit{~} MDS(0,1)$
$$ \sigma_t^2 = w + \alpha_1 y_{t-1}^2 + \alpha_2 y_{t-2}^2 $$
I am trying ...
0
votes
1answer
21 views
What exactly is meant by “constant autocorrelation structure” in the definition of stationarity of a time series?
I have come across the term "constant autocorrelation structure" in every definition of stationarity but I have not been able to find an explanation of this phrase. I know that ...
1
vote
2answers
22 views
What is the lag associated with Moving Average smoothing?
In a tutorial I came across this:
"Recall that the forecast value is: $\hat{y}_{t+1} = \frac{y_t + y_{t-1} + ... + y_{t-m+1}}{m}$
It's worth pondering that formula for a minute. While easy to ...
0
votes
1answer
34 views
Time Series Analysis vs Linear Regression for Stock Price Prediction
I see many online blogs that use Linear Regression to predict future stock price.
I too have done this and my x variable is time elapsed.
But I have been advised this problem is better suited for Time ...
0
votes
0answers
12 views
Time series-Variance of Y-bar
https://imgur.com/a/YTn4zYC
Hello, I currently have the solution for a Time series variance problem I am facing . I am aware that the U (mu) is a constant and will go to zero , the 2 error terms are ...
0
votes
0answers
10 views
How can I measure time to effect using the Causal Impact package? [closed]
I am delivering a drug and measuring a response variable. Using the response variable (a single vector vs. time) I am trying to measure (1) if there is an effect caused by the intervention (the ...
0
votes
0answers
11 views
Assess timeseries sample representativeness
Problem
I want to build a benchmark (for forecasting models/algorithms/auto model selection) that runs over a subset of all my timeseries (each timeseries is a different case/business).
Question
Is ...
0
votes
0answers
5 views
Should we log transform non-zero time series data or let the automated methods find the right value?
I have data that contains only non-zero values - deaths. The data is heteroscedastic hence I used lambda=0 (log transformed it). When I use ...
0
votes
0answers
3 views
Data leakage when setting class_weight to tackle imbalanced time series data?
I'm using a random forest classifier from sklearn to predict whether a stock's return for the next period is greater than a certain threshold (say -2%), so negative is 0 and positive is 1, a binary ...
3
votes
0answers
30 views
+50
References to time series feature stability with varying series lengths
Firstly I'd like to note that according to a discussion in meta this question is indeed on topic.
I have trouble finding relevant research related to time series analysis. I'm interested in how ...
0
votes
1answer
12 views
Fitting an ARIMA subset model in R
Suppose I have a time series. After looking at the PACF plot, it largely decreases to zero after 3 lags, but there is also a PACF value that "pokes out" of the significant bounds at a far ...
2
votes
0answers
22 views
Searching for Clive Granger's 1972 paper on time series decomposition from the Budapest Conference
In 1972 Clive Granger presented a seminal work on the Box-Jenkins Time Series methods as a conference paper:
C. W. J. GRANGER, Time Series Modelling and Interpretation, Paper presented to the European ...
0
votes
1answer
13 views
Bivariate cointegration in a multiple regression error correction model
I am new to cointegration and ECM.
I have two I(1) variables that I have estimated and their linear combination is I(0) as per the Engle-Granger test.
Is it then possible to use this error-correcting ...
1
vote
0answers
5 views
Fitting ARMA-GARCH with t-distributed error
I am currently trying to find parameter estimates to a ARMA-GARCH model with student's t distributed innovations.
However, the errors are t-distributed so there are no closed form MLE. As for finding ...
0
votes
0answers
16 views
How do I decide in which sequence I add my regressors to the ARIMA model?
I have time series data (one DV and 26 IVs) and want to estimate ARIMA models with multiple regressors in R. I have one time differenced my DV, so d=1. With the help of the ACF plot and the auto.arima ...
1
vote
1answer
15 views
Changes in correlation between two variables in VAR model after differencing
I am building a VAR model with two variables.
When I used it with non-stationary data with strong trend and seasonality, the correlation between variables was 0.48.
...
0
votes
0answers
11 views
Producing 'structurally reconciled' forecasts (e.g. revenue = price * volume)
I recently learned about Hierarchical Forecasting and that this principle can be applied across hierarchies, cross-sectional groups or temporal aggregations. Independent Forecasts are created and the ...
0
votes
0answers
9 views
In case of “No breakpoint estimated”, should I use only linear regression instead of time segmented analysis in R?
I am trying to do 2-time segmented analysis using the package segmented in R but I am encountering errors that I tried to figure ...
0
votes
0answers
10 views
LSTM - unsure about input shape (dataset with multiple instances of patients)
I read through a lot of similar questions about the input of LSTM's, but since the applications were different from mine and mostly used one dimensional data it confused me more than it cleared ...
0
votes
0answers
21 views
Random walks and Pearson's correlation
One well known problem in time series analysis is spurious correlation when time series are non-stationary (and non-cointegrated). Given random walks of the form $R_i=R_{i-1}+\epsilon_i$, with $\...
0
votes
0answers
12 views
Predicting time series for multiple clients
I have a dataset with 100 people that are selling products per week.
The data in the image is not real because it is confidential, but the data looks exactly like this. The only thing to mention is ...
0
votes
1answer
34 views
Time series analysis: multiplicative model and seasonal adjustment of data
I am trying to help a friend in statistics and this question involving time series came up and I did not know what to do. I tried searching different stack exchange forums for answers, but I believe ...
0
votes
1answer
22 views
Is it possible we supply the differenced data to the ARMA model (not ARIMA) and have the model forecast for the original time series?
I am trying to fit an ARMA model in Python for a non-stationary series. Conceptually, is it possible that I difference the data to first convert it to stationary data, fit the model on this ...
1
vote
0answers
19 views
Forecast horizon and small sample problem
I have a stationary time series with slight persistence, if any; like stock returns. They are collected in weekly frequency, $N$ obs. I try to predict them with an $AR(1)$ model. The well known ...
0
votes
0answers
17 views
CausalImpact - Choose the predictor time series
I am interested in the CausalImpact to analyse the impact of an event on a time series. Refer also to:
https://storage.googleapis.com/pub-tools-public-publication-data/pdf/41854.pdf
https://google....
0
votes
0answers
13 views
Time Series Predict Frequencies or Times
I have a multi scale stochastic time series where I am forecasting an integrated future. I was wondering if I should focus on predicting the series at different time intervals in the future or at ...
1
vote
1answer
19 views
Would serial correlation be problematic if only the regressors are serially correlated in a multiple regression model?
In this regression:
$Y_t=\beta_0+\beta_1 X_{1t}+ \beta_2 X_{2t} + ... + \beta_p X_{pt} + \epsilon_t$
If there is serial correlation among e.g. $X_{1t}$, what would be the consequence?
0
votes
0answers
16 views
How to predict a discreet outcome from time series?
I was asked (not homework) to produce an inferential model where:
independent variables are:
vital signs (13e6 samples, 5 time series variables, + continuous values, NMAR)
hemoglobin values (20000 ...
0
votes
0answers
18 views
Anomaly detection for multiple correlated variables (sensors)
I have time series data from multiple sensors detecting vehicles driving through the sensor zone. I'm aggregating the total actuations for each sensor in 15-minute bins. The sensors are correlated, ...
1
vote
0answers
42 views
Can an MA(1) process arise from a different MA(1) process plus white noise? If so, what are the constraints on the parameters of the two processes?
This is a question linked to an earlier discussion (What subset of ARMA(1,1) processes can be represented as AR(1) - a query about the logic in this derivation) .
I am looking at the possibility of ...
0
votes
0answers
26 views
Hyperparameter Tuning with Keras / Tensorflow for multivariate time series regression
The model
I am training a dense feed-forward NN using the Keras API on Tensorflow.
Each sample of the training set defines $\mathbf{X_t}$ and $\mathbf{Y_t}$ of an observed time period $t\in T$. Input ...
0
votes
0answers
35 views
Choice of Time Series Models for Forecasting Data
I have a question that to most readers here probably think is fairly simple to answer, but never the less, out of curiosity I'd like to ask:
Given a time series, yt, which is stationary (say the log ...
0
votes
0answers
6 views
Testing for MCAR in univariate time-series
I have a large dataset consisting of hourly energy measurements over a period of 2 years for 1500 buildings. Since the energy consumption of one building can be assumed to be independent of the ...
0
votes
0answers
7 views
Predicting Multiple Values Values Using Time Series Forecasting
I want to illustrate my question with the following example:
I have a wholesale company through which I sell 200 products:
P1,P2,P3 .... P200
to a 1000 customers ...
0
votes
0answers
18 views
I am modelling a time series, the results are coming fine without exogenous variables but predictions are getting wrong on including exog varaibles
I am modelling a time series, the results are coming fine without exogenous variables but predictions are getting wrong on including exog varaibles, the forecasting series is starting from a lower ...
0
votes
1answer
19 views
What is a good auto-correlation value?
I have a dataset of 400 time series. I want to evaluate if there is some relationship between consecutive data points. Thus, I have calculated the auto-correlation (AC) of the time series with ...
0
votes
0answers
9 views
Theoretical ARIMA Model From Quadratic Series
I'm trying to find the form of an ARIMA model from the model, $y_t = \beta_0 +\beta_1 t +\beta_2 t^2 +\epsilon_t$ where $\epsilon_t $~ $WN(0,\sigma^2)$
I started by differencing the series:
$\beta_0 +\...
0
votes
0answers
26 views
Using visual representation of time series of unequal length
I would like to apply methods like Gramian angular field, recurrence plots and Markov transition fields to a time series classification (TSC) problem where the time series themselves are all of ...
0
votes
0answers
14 views
Questions about distributed lag non-linear models (DLNM)
I am new to DLNM, I have some confused questions:
My research is about the environmental impact on health, and my data records are ordered as health cases as below:
I want to use DLNM to analyze the ...
