Time series are data observed over time (either in continuous time or at discrete time periods).

learn more… | top users | synonyms

1
vote
0answers
14 views

Fit parametric distribution to time-series data

I have a time-series of daily revenue data for a number of years and I would like to perform a parametric distribution fit to the data. Given that distribution fitting requires i.i.d. data, I presume ...
1
vote
0answers
20 views

Arima time series

I am trying to build a arima time series model for demand prediction... my data is on weekly level from 2014 and 2015 all weeks. If I also use 2016 first 10 weeks data and then try to predict the for ...
0
votes
0answers
25 views

How to manually predict one step ahead time series data using coefficientes estimated by arima function in R

My objective it to manually compute one-step ahead forecast using the estimated coefficientes given by the arima function in R. I will consider the specific model ...
0
votes
1answer
45 views

How much correlation is significant enough to build a regression model?

I am trying to build a regression model using two time series data in R. There is not much correlation between the two time series, so I am using trend part of both time series(using STL decomposition)...
0
votes
0answers
43 views

Stationarity of time series and VAR model

I have two variables, one is stationary I(0) and one is non-stationary I(1). Is it possible to make VAR model for these two variables if the non-stationary variable will be differenced to obtain ...
1
vote
0answers
12 views

Mixing types of Interpolation while Displaying Data

We have a piece of in-house software for pulling data out of our system and displaying it. It can do this at the resolution of the data, aggregate it up to a lower resolution or down to a higher ...
1
vote
1answer
10 views

Detecting trend divergence between two time-series datasets

Is there a relatively simple to implement method out there for detecting the early stages of divergence between two time series ? To take a simple example from the financial area, a common one is ...
1
vote
0answers
25 views

Determine best ARIMA model with AICc and RMSE

I have done a training set to fit different ARIMA models and then a test set to assess their performance (with R). From what I understood, I can use the AICc to determine the best model by choosing ...
0
votes
0answers
19 views

Sales forecasting: Unsure about data grouping

I am trying to implement a simple, short-term (1-4 weeks) forecast of product revenue/sales. The data I have is brand category product revenue where ...
0
votes
1answer
25 views

What imputation methods can be used for missing not at random covariate values in a survival analysis?

I'm new to survival analysis and trying to understand how to use it properly. My dataset is a time series dataset where most dependent variable values are available, 2 dependent variable values are ...
42
votes
9answers
6k views

What is wrong with extrapolation?

I remember sitting in stats courses as an undergrad hearing about why extrapolation was a bad idea. Furthermore, there are a variety of sources online which comment on this. There's also a mention of ...
0
votes
0answers
18 views

Correlation between yearly store-sales and cross-sectional time series

I have a monthly time series data (for five years) on a particular store about its sales, marginal sales, inventory, etc. I also have monthly cross-sectional data (for five years) on the same. I ...
0
votes
0answers
9 views

Grouping target variable into bins of minimal variance for uniformly-observed, seasonal timeseries data

I am looking at a key performance indicator that is measured uniformly over time for which I strongly suspect seasonality. I would like to create groups to identify periods of time where observations ...
0
votes
0answers
19 views

Disadvantageous of non stationary process

If the process is not stable, as i konw the prediction will be impossible and the mean and variance will be instable or even infinite, and may be there is a correlation between variables. Are there ...
0
votes
0answers
15 views

Innovations algorithm in Matlab [migrated]

This seems like it's supposed to be a simple task but going through the "predict" documentation in Matlab I found this to look unnecessarily complicated so I could be looking in the wrong direction. ...
2
votes
0answers
19 views

Changing sensitivity (cval) in tsoutliers resulting in unexpected results

I am using the excellent tsoutliers R package to detect outliers (additive outliers, temporary changes etc.), but the cval ...
0
votes
0answers
14 views

Price Prediction for more multiple products

I need to arrive at a guidance value for multiple products (more than 10K) for multiple customers. I am building two segmentation models - customer and product segmentation to suggest the price. On ...
1
vote
0answers
10 views

Method for analyzing interactions of periodical exposures (drug x drug interaction and likelihood of adverse events)

I am interested in studying drug usage interactions (drug x drug) and its relation to adverse events. I would appreciate help with choosing the right model/method for this type of analysis. The data ...
0
votes
0answers
13 views
0
votes
0answers
21 views

GARCH Modeling Production time series

I'm working on a factory's production data (chronological) and I want to apply GARCH-ARMA Model on it, here is the plot of the data. Is applying GARCH-ARMA a good idea? If so, how do we come to the ...
0
votes
0answers
14 views

Non-parametric smoothing a small sample time series with fixed/known t(0) and t(n)

I would be grateful for any suggestions on how to smooth a time series with the following properties: We observe $t(i)$ for every integer $i=0...T$. $0 < t(i) < \infty$ $T$ is typically small (...
0
votes
1answer
19 views

Time Series : Proof of the Moving Average of order 1 model

For those of you who are versed in time series theory, can you help me with this little question ? Can someone explain to me the concept behind the theta and it's properties? Is it a constant ? and ...
0
votes
0answers
13 views

Relation discovery between two time series data

I'm looking at analyzing the relation between temperature & sales/searhes of particular product at a daily grain The relationship is little complex, for example Sales go up for low temperature ...
0
votes
0answers
16 views

State space modelling of longitudinal data in r

I have n stations, and for each station there are m time series observations on different days, each of length ...
0
votes
0answers
11 views

How to measure the impact of a product on the global revenue?

We have products daily revenue and when each product started being sold by the company. We also have different break-down of product revenue by location, store, sales person, etc. We need to measure ...
0
votes
0answers
27 views

Maximum likelihood estimation for non-stationary time series

I want to estimate how the taxes influence the retail price of alcoholic beverages. The price function is tricky because in EU countries there is excise duty and also VAT. The non-linearity (which is ...
3
votes
1answer
25 views

Is weakly stationary equivalent to $I(0)$?

I'm currently reading some time series lecture notes. It says that: Weakly stationary (or wide-sense stationary) processes are said to be $I(0)$ (integrated of order $0$). Let's call the above ...
0
votes
1answer
21 views

Simulating a Stochastic Integral of OU process

The stochastic integral I want to simulate is $$\int_{0}^{1}J_c(s)dJ_c(s)$$ where $J_c(s) = \int_{0}^{s}e^{-c(s-r)}dB(r)$, is an OU process. I simulate the data using Matlab and the sample codes are ...
1
vote
0answers
15 views

Cycle detection of a time series graph (y axis power) - machine productive cycle vs idle times

I'm fairly new to data analysis so I'm not exactly sure on how to approach this problem. I have obtained data output from an industrial machine which has 2 states - productive and idle. The productive ...
0
votes
1answer
43 views

adf.test vs ndiffs

I have a really small time series dataset (21 yearly observations) and I want to check if my data is stationary. ndiffs(TS, test="adf") ...
0
votes
0answers
19 views

Statistical analysis of two groups of time series

I am working on an evaluation framework for algorithms applied to video sequences but I'm stuck on how to correctly do the actual evaluation. I'll try to explain my data as best I can, however, my ...
1
vote
0answers
42 views

LSTM mimicking unseen time series data during testing

I have built a LSTM network which has been trained on a time series dataset (which is week-wise logged). The LSTM is able to make pretty accurate predictions as of now. Training data seems to have ...
1
vote
0answers
23 views

Reverse-Engineer Time Series Matrix With Machine Learning

I'm trying to figure out the following situation which is almost the same as in this post here Time series with multiple subjects and multiple variables ...
1
vote
0answers
23 views

GMM Moment Condition

I have encountered some difficulties in solving the following exercise: Consider $(y_{i},x_{i})$, which are i.i.d pairs. We observe a random sample from this distribution. We want to estimate $\mu=E[...
0
votes
0answers
41 views

Adaptive baseline for time series data for anomaly detection

I would like to create/calculate a dynamic baseline for continues time-series data. The data is arriving in real-time (streaming) every N minutes interval. I googled around and came across control-...
0
votes
1answer
93 views

Lagging/Leading Indicator Length Time

I tried looking this question up on google and didn't find material that answered my question. But my questions are: (1) Is there a method to determine how long it takes a leading indicator to ...
1
vote
1answer
20 views

Example of 2 series correlated but not cointegrated and vice versa

I am studying the time series and only kind of understand correlation vs cointegration. Can someone provide an example of two series that are correlated but not cointegrated, and two that are ...
0
votes
1answer
37 views

Intuitive explanation of state space models

Having looked into options for modelling and forecasting a financial time series based on mixed frequency data, I came across state space models, which seems worth exploring. I've however been ...
0
votes
0answers
11 views

choosing test set in a seasonal time series

The data is sequential, but not necessarily continuous, ie. there are multiple gaps between the start and end date. I fit a regression model, which may or may not involve lagged variables, and I want ...
3
votes
0answers
77 views

How to select the length of a time series when fitting models for prediction

Let say that one wants to fit a model to a daily financial time series for prediction (e.g. ARIMA, SVM). If data are stationary, ideally the longer the time series, the better. In practice, I don't ...
0
votes
1answer
28 views

When is time treated as a discrete variable?

Time is usually treated as a continuous variable but in some cases it is discrete. An example would be with a drug study and measurements are taken at 1, 2 and 3 hours. Am I right to think an ...
0
votes
0answers
13 views

Repeated measures MANOVA for testing difference among multivariate time series?

I measured behavioural multivariate time series (3 variables) from 7 dyads performing an experimental task. For each dyad, I have a multivariate time series (that is, I don't have data from each ...
1
vote
0answers
29 views

Can Random Forest regression handle non-stationary input variables?

I am working on a project where the explanatory variables include soil attributes, land use and land cover properties, stream flow and climate (precipitation, temperature etc) measurements recorded at ...
1
vote
1answer
39 views

How to correct a variable for unit root?

So I have done my dickey fuller test on the variable exchange rate. The result table indicated that I do have unit. In addition, I have tried lagging the variable but still ended with a result that ...
1
vote
1answer
30 views

When would the data set be stationary?

I am not a statistician or mathematician and I need some help. I did three experiments as I labeled on the figure, 3,4,5, each experiment has x and y results. The markers are the real data and the ...
0
votes
0answers
3 views

How to plot the forecasted values against actual values observed later in R? [migrated]

We used the R library forecast to make predictions for the next 24 hours. We have the following: ...
0
votes
0answers
32 views

Forecasting with mixed frequency data

Just a general question that I couldn't find too much on. What would be some good approaches to one step ahead forecasting of financial time series with mixed frequencies? Often a lot of the ...
1
vote
1answer
33 views

Why unit root test with D-F not using normal or t test?

Let $X_1, \dots , X_n$ be observations from the AR(1) model. For large $n$, the maximum likelihood estimator $M(\phi_1)$ of $\phi_1$ is approximately normally distributed as $N(\phi_1, (1−\phi_2)/n)$...
2
votes
1answer
24 views

What is the approporiate test?

I have two variables ; First variable is ordinal and it represent years series from 2006 to 2015 and another variable which is interview score and I want to study if there a relationship between the ...