Questions tagged [forecasting]

Prediction of the future events. It is a special case of [prediction], in the context of [time-series].

1,003 questions with no upvoted or accepted answers
Filter by
Sorted by
Tagged with
1
vote
1answer
31 views

Does forecasting with ARIMA lose non-stationary components?

Suppose I have a time series $Y$. I have read that an ARIMA model consists as an ARMA model of a stationarized version of $Y$. If I try to predict $n$ ticks ahead with an ARIMA forecast model (with $...
1
vote
1answer
36 views

Is normality test a must for SARIMA forecasting?

My data is not normal and I have tried Box-Cox transformation, yet after Box-Cox transformation, it still fails under the Kolmogorov-Smirnov test, so can I skip the normality transformation and use ...
1
vote
0answers
18 views

Dependent forecasts from multiple models

I have n forecast models that are run independently but I also want the value of the forecasts to sum to 1 and each forecast to be >= 0 in order to preclude some arbitrage conditions. I’d be ...
1
vote
0answers
129 views

Evaluate the conditional variance forecast from a GARCH model

I wanna evaluate a simple GARCH(1,1) model for the conditional variance. Firstly, I understand that the conditional variance is unobserved and that is really the crux of the issue. Out-of-sample, I ...
1
vote
0answers
226 views

How to predict disk usage behaviour?

I've recently been tasked with developing a prediction model for the scientific computing cluster at my university, with thousands of users. User behaviour typically looks like this: Most users are ...
1
vote
0answers
77 views

multi-response forecasting

General Dear community, I really struggle with some imporant issues for my next project. In general, the investigation is about multi-response forecasting with financial data. The predicability of ...
1
vote
0answers
321 views

Forecasting with ARMAX vs. Regression with ARMA errors

In this post Rob Hyndman says that for forecasting, it doesn't matter whether we fit an ARMAX model or an OLS model with ARMA errors: https://robjhyndman.com/hyndsight/arimax/ Why is that the case? ...
1
vote
0answers
92 views

Fourier Output Meaning

I just ran a fourier series on weekly sales data for 3 years worth of data. I optimally chose the number of k-terms based on the AIC. First 6 lines of my data: ...
1
vote
1answer
87 views

Need advice concering forecasting next year based on irregular time-series - UPDATED

I need your advice with regards to the following inquiry: "Based on your observations, what could you say about the load for the same months in year 2019?" ...
1
vote
1answer
104 views

Updating a hypothesis on multiple partitions of uncertain evidence

I want to forecast $P(A)$ where $A$ is a messy real-world event, for which I have no analytical expression or statistical model. Assume, however, that for $b$ events $B_i$ I have forecasts for $P(...
1
vote
1answer
36 views

Sales Forecasting for Multiple Dealers

I have a data-set containing 7046 unique dealer codes and their monthly sales data from April 2013-August 2018. The Financial Year for the sales data begins in the month of April for a year and ends ...
1
vote
0answers
32 views

Performance of Hierarchical Temporal Memory on unsupervised online anomaly detection problems

I'm looking for an algorithm to detect anomalies in streaming data (server metrics). The detection needs to be near-real time and unsupervised (labeled data will never be available, unfortunately, and ...
1
vote
0answers
132 views

Prediction of time series with neural network

Suppose I get a forecast, from MLP or LSTM - next 7 time steps into the future. I can assess its quality using mean absolute error using cross validation. However, it is not clear, what I should do ...
1
vote
0answers
112 views

Intuition about Exponential Smoothing parameters?

If I use Triple Exponential Smoothing with Additive Seasonality and let a statistical program optimize alpha, beta and gamma for me, is there something I can conclude about my data based on the ...
1
vote
0answers
27 views

Time series analysis where computing the exact value is possible (but expensive)

I have a stationary time series where it is actually possible for me to compute the exact next value. These computations are very expensive, and to speed things up I want to employ following scheme: ...
1
vote
0answers
33 views

Multivariant time series forcasting

I have a multivariate time series data with Timestamps as( 'season' with four categories: 1,2,3,4, 'month' with 12 categories and 'day' with 5 categories excluding Saturday and Sunday) and I have to ...
1
vote
0answers
49 views

Johansen test R - long term relationship covariance

I have two time series which are I(1) and co-integrated. I would like to make long term forecasts for one of the time series, given an assumed fixed value for the other. I used the Johansen test (ca....
1
vote
1answer
30 views

Macroeconomic effects. Effects of a time serie on another

I have a monthly time series for the provision in a financial institution. Take real data until december 2017 and predict it with a Bat model until June 2018 using R and I have an error of 0.12%. This ...
1
vote
1answer
45 views

Forecasting new period with Regression Model

Have a dataset which consists of item names (rows) and monthly values of sales (columns). My task is to predict value of sales for next month and I'm trying to use regression models for that. But the ...
1
vote
2answers
89 views

Forecasting sales: different methods

It's a bit untypical question I guess, but I hope you can help me. I know a little bit of statistics. I'm not a specialist, but I find it really interesting, so I learn in my free time. I need your ...
1
vote
2answers
382 views

Dealing with missing data in Time Series or non-constant time intervals for forecasting in R (ARIMA, Holt Winters, Theta)

I have a time series of sensor data from a machine. This machine is sometimes moved and thus there are big chunks of missing data, here is a plot of the data points: My goal is to try to start ...
1
vote
1answer
139 views

In prediction, when should I use rolling windows vs. nonoverlapping ones?

Suppose I have daily time series data and I want to predict a month in advance using a set of features. I have lots of them so I'll be using regularized linear regression. To create the response I can ...
1
vote
0answers
80 views

Autocorrelation in a multiple regression time series model

I am developing a multiple regression model that uses inputs available at time t to predict a continuous outcome at time t + 1. One of the best predictors of the outcome at t + 1 is the value for the ...
1
vote
0answers
394 views

Not available with this estimation method (ARMA ML) error message in eviews

I tried a Breusch–Godfrey LM test on an ARMA Maximum Likelihood model in eviews and I got the following error message: Not available with this estimation method (ARMA ML). I re-estimated the equation ...
1
vote
0answers
112 views

Measuring impact of intervention using a counterfactual forecast?

I'm an economics student who's trying to dive into causality. I've tried looking into intervention analysis, but I can't seem to find many sources on forecasting using historical data as an approach. ...
1
vote
1answer
88 views

How to choose the number of steps ahead when comparing time series CV to the AIC or the BIC?

I would like to empirically evaluate the performance of the AIC, BIC and Cross Validation as model selection criteria for time series forecasting, i.e. which one of these criteria leads to the best ...
1
vote
0answers
43 views

Weekly forecasts when some predictors are monthly

I would like to forecast at the weekly level, but some of the predictors are at the monthly level. The only way I can think to approach this problem is to divide the monthly total by the number of ...
1
vote
0answers
14 views

AR on Non-Linear Time Serires

Should autoregressive model be able to fit and forecast non-linear time series, for example, a sine curve (by non-linear I mean the data generating process)? I used ...
1
vote
0answers
389 views

Using a log transformation in an ARIMA model; how to get model predictions back on original scale?

This question relates largely to the answer provided by Stephan here: How to know which forecasting tool to use in R? Stephan says: If you work on logged data, be aware that simply taking the ...
1
vote
0answers
199 views

Diebold-Mariano Test (Newey–West adjusted)

I've two questions regarding to the Diebold-Mariano test in comparing predictive acuraccy. I am reading this paper here. Differences in MSPE are reported together with p-values from the Diebold– ...
1
vote
1answer
23 views

Parameter estimators of linear predictors

Suppose a linear predictor of the form $a + b'X$. To find estimators for a and b, should we minimize $E[Y-a-b'X]^2$ or $E[(Y-a-b'X)^2|X]$. Former gives $\hat{a} = E[Y] - b'E[X]$ and latter gives $\...
1
vote
0answers
15 views

Is there any correlation between forecasting errors from forecasting done at different origins?

Let $\ e_{T+l|T} = Z_{T+l} - \hat{Z}_{T}(l)$ be the forecasting error $\ l$-steps ahead when the forecasting origin is time $\ T$. Now, let $\ e_{T+l-j|T} = Z_{T+l-j} - \hat{Z}_{T-j}(l)$ be the ...
1
vote
0answers
24 views

Can you do demand/sales forecasting using data with no significant correlation?

I am playing around with a sales data period January 2016 - July 2017 from a company, after using ACF and PACF, this is the result: ACF: PACF: I am currently trying to do a sales forecasting based ...
1
vote
0answers
185 views

EWMA using Monte-Carlo simulation

Im trying to forecast volatility using an EWMA model in python. Where i have return(t-1) and variance(t-1). n is number of days. for every Monte-carlo simulation N: t=1: Forecast the variance using: ...
1
vote
0answers
72 views

Forecasting turnover on retail industries with time series

I'm working on a project of forecasting turnover on retail industry. I have the turnover of different products for a 2 years period of time My final goal is to forecast a global turnover with and ...
1
vote
0answers
32 views

StatisticLaw behind air traffic demand

I am performing linear regression in order to see the relationship between Air traffic demand, GDP per capita and fares of ticket. I am interesting in getting elasticities therefore, I used the ...
1
vote
0answers
87 views

Forecasting formula in mathematical notation help please:

Question: is what I have correct? If not what is correct? I am not sure how to correctly write this forecasting model in proper mathematical notation. My attempt is at the bottom, please help fix or ...
1
vote
0answers
122 views

Can i use Diebold Mariano test for comparison of 2 models across multiple time series?

I have 2 models (for simplicity, let's call them AR(1) and MA(1)) each making 1 day ahead forecasts of time series. If I had only 1 time series I would just use ...
1
vote
0answers
378 views

Can Croston's method handle a series that has a large number of zeros at the end?

I am trying to implement Croston's method in Python. The way I understand it, the first step is to take a series of the form: ...
1
vote
0answers
224 views

When optimizing $\alpha$ in a simple exponential smoothing model, is there any benefit to using something more sophisticated than least squares?

I am trying to manually implement simple exponential smoothing, for which the formula is pretty straightforward: $\hat{Y}_{t+1} = \alpha Y + (1- \alpha) \hat{Y}_t$ In the original formulation, the ...
1
vote
0answers
255 views

Forecasting with ARIMA models and sensor data in R

I'm trying to apply ARIMA models to sensor data and would be thankful if anyone could answer my questions. I should add that I have very little experience with time series (trying to change that). ...
1
vote
0answers
178 views

How to make multiple step ahead predictions with cv.glmnet object?

I am trying to make forecasts for a LASSO model obtained from the cv.glmnet() function ("glmnet" package). I most frequently make forecasts using the predict() function (in the "stats" package). For ...
1
vote
0answers
160 views

A forecast distribution implementation for Croston's model?

I read the paper by Shenstone and Hyndman on the distributional forecast underlying Croston's model. I looked into the forecast R package and the tsintermittent R package. Neither seem to have the ...
1
vote
0answers
2k views

No suitable ARIMA model found error

I am trying to forecast the sales for next 48 days from the data given by modelling for multiple seasonality and day of week , promotional effects. R could not come up with a suitable model. I need ...
1
vote
0answers
51 views

Load forecasting using LSTM

I am using MATLAB 2018a, and looking for an explanation, or pointers, on how can I modify the matlab example for "sequence to sequence regression" to use it with "double" type predictor data array ...
1
vote
0answers
28 views

Shall my variable be differentiated when I forecast? (EViews)

I've read this http://www.eviews.com/help/helpintro.html#page/content/Forecast-Forecast_Basics.html before and have apply some certain things in EViews (student version, btw) and I would like to ...
1
vote
0answers
58 views

Predictive regression on Fama/French

I have to predict (monthly) returns on Stock Indices (S&P 500) with the FF-Model (3 and 5 Factors). Therefore I shall use a predictive regression and an in-sample analysis. I started off with a (...
1
vote
0answers
368 views

What is the best way to add lagged features for time-series predictions with a long prediction period?

I am currently dealing with a time-series problem and want to add lag feature(s) to my data. The point is, that I should predict 1 month of daily sales based on 4 months of training data. So if I add ...
1
vote
0answers
9 views

Index creation for studying “average” time series properties

I have 300+ financial time series from an "unknown" asset class. To study the dynamic properties of class my idea was to collect them under an "index" and then study it as an univariate time series. ...
1
vote
0answers
26 views

Time Series Forecasting using Custom Representation

Let's assume that I have a background knowledge of some stochastic nondeterministic process and I identified the process to be autoregressive and follow some irregular pattern: $z_t = \sum_{i=1}^N \...