Questions tagged [forecasting]

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

1,031 questions with no upvoted or accepted answers
Filter by
Sorted by
Tagged with
3
votes
0answers
134 views

Does Nickell bias matter in forecasting?

The context is longitudinal data, with $i$ indexing individuals and $t$ indexing time. The goal is predicting $y_{it}$ as a function of lags of $y$ as well as $\mathbf{X}$, which might include lags. ...
3
votes
3answers
94 views

Appropriateness of time series regression for intensive longitudinal data

I am analyzing time series data in which participants rated their thoughts in real time. I am trying to model the shape of the data. Details on the time series: Sampling rate was 1/4 second ...
3
votes
0answers
139 views

LASSO: k-fold Cross Validation of AR(p)-process

To improve my intuition on shrinkage models, I want to "recode" the lasso by myself. However, I'm at the point, where I have to program the k-fold Cross-Validation. At my future application of the ...
3
votes
1answer
507 views

Trending time series data normalization for Deep Learning

I'm replicating following article Financial Time Series Prediction using Deep Learning and I'm stuck with data normalization. In chapter 5.1 in the second paragraph in the last sentense the authors ...
3
votes
0answers
295 views

Assumptions on Neural Networks (NNETAR)

Are there any assumptions that must be covered when fitting an NNETAR model? non-correlation, normality, or something? I've already saw Rob Hyndman post where he says NNETAR doesn´t need stationarity, ...
3
votes
0answers
1k views

How does neural network auto-regression produce multistep forecasts?

I am looking at time series forecasting using neural networks as described in Hyndman and Athanasopoulos. They describe Neural Network Auto-Regression models as non-linear generalizations of AR ...
3
votes
0answers
211 views

How to interpreate Reliability Diagrams?

In general the interpretation of reliability diagrams isn't a problem for me, to identify if a forecast is over- or underestimated. But in this dataset which contains different product probabilities ...
3
votes
1answer
80 views

How to account for remainder in forecasting?

I've done an STL decomposition of a time series. While trend accounts for major variations, remainder has a "cyclical" setup. I want to know what other forecasting methods I can apply to maximize ...
3
votes
0answers
244 views

How Negative Binomial Distribution and negative bionomial regression can be used to sales forecast?

My first question here. Due to the improper inventory management we seem to have dispersed sales, and the stores are unable to meet the demand because items are being out of stock. There are so much ...
3
votes
0answers
153 views

Steps to find optimal transformation for wide-sense stationarity

I've been trying to automate the procedure of choosing the best transformation for a non-stationary process (in R). For lack of a better term, "best transformation" here refers to the quality of ...
3
votes
0answers
61 views

Is 5 in 100,000 the same as 50 in 1000,000 from a Bayesian perspective?

I am learning about bayesian reasoning and I was having a discussion with a friend about it. My friend asked me this question, which I couldn't answer (this is not a homework, I am an adult learning ...
3
votes
0answers
43 views

How to include handle days with no delivery?

I am currently trying to forecast the delivery of cash to branches. The problem I face is that there are a few days, as well as most Sundays, in which there are no deliveries. Hence, in the time ...
3
votes
0answers
119 views

AIC comparison vs. unit root test for model selection in forecasting

Rob J. Hyndman once wrote in "Why I don't like statistical tests" (emphasis is mine): In forecasting, the only place in which I find testing useful is in determining the order of integration of a ...
3
votes
0answers
2k views

How to estimate weekly and daily seasonality for data with 15min frequency in Python?

I am relatively new to time series. My goal is to predict a few hours of data, measured every 15min based on three months of observations in Python. I assume I have daily and weekly cycles which I ...
3
votes
0answers
142 views

Holt-Winters: Can I use more than one seasonal cycle for SSE minimisation?

I am minimising SSE to estimate the parameters and starting values for a Holt-Winters model. I.e. "forecasting" the values using different parameters, measuring the sum of squared errors of these "...
3
votes
0answers
186 views

Predictability of predictor variables in regression analysis

I have run a multiple linear regression analysis to predict the forecast of demand (in litres) of soft drinks. I have 104 sets of weekly data and my independent variables are feature space (measured ...
3
votes
0answers
100 views

Is there a general and completely automated (i.e. programmed) solution to modelling and forecasting time series data?

Is there a general and completely automated solution to modelling and forecasting time series data? I think this question is extremely important. If it is not possible, please provide an ...
3
votes
0answers
25 views

How to make shrinked forecast for the extreme value?

Let me use made-up example: John loves running. He decided to run in his local half-marathon for the first time in his life. He never measured exactly how fast he runs the distance, but while ...
3
votes
0answers
167 views

VAR model with different time period for each series

I am trying to fit a Vector Autoregression model to forecast GDP growth Rate. I have 2 series, monthly GDP growth rate and a monthly economic indicator. For the monthly GDP growth rate, the latest ...
3
votes
0answers
893 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 ...
3
votes
0answers
133 views

Comparing variances of forecast errors

I am forecasting a weekly commodity price series. I use a rolling window for estimating my model, and from each window I make point forecasts for one and two steps ahead. I want to investigate ...
3
votes
0answers
572 views

What's an example of a best linear predictor that's not a best predictor?

I just learned about the definition of a best linear predictor found by minimization of variance of $Y$ given $X$, or in other words trying to minimize the variance of $$\mathbb{E}\left(\{Y-\hat{Y}(...
3
votes
0answers
292 views

Computing Seasonality Index and application to shorter time series

Background: I have an overall time series of close to 3 years of data. I need to forecast for different slices of data. When I slice the data, some slices results in a shorter time series. We go with ...
3
votes
0answers
336 views

How to calculate prediction intervals based on Chebyshev inequality?

I have recently read the article by Gardner (1988) who proposes Chebyshev inequality-based prediction intervals for forecast: suppose we have a model selected on the usual basis of one-step-ahead ...
3
votes
0answers
315 views

is it sensible to use monte carlo to predict sum of time-series over an interval?

I have created a model that forecasts out a time series at the daily level along with prediction intervals two months into the future. There is little to no auto-correlation in the time series so I ...
3
votes
2answers
125 views

Evaluating probabilistic forecasts of K-most-likely events from an arbitrarily large event space

Suppose a populous nation has a high homicide rate and an understaffed police force. The police chief hires a statistician and together they decide to take a preventative approach by identifying ...
3
votes
0answers
180 views

Prediction intervals for forecasts using spectral analysis

I have circadian data which typically have a period of around 24 hours so using spectral analysis seems appropriate. I've used spectrum resampling which is quite robust to changes in period which ...
3
votes
0answers
226 views

Forecasting demand with out-of-stock data

Usually retailers have a service level that is below 1.0, which means that share of products is out-of-stock some of the time. What is the best practice or possible ways of using out-of-stock data to ...
3
votes
0answers
166 views

Reasons for GLM ('identity') performing better than GLM ('gamma') for predicting a gamma distributed variable?

I am investigating different methods for fitting my target variable (observed wind speed: positive, real, with small values being most probable) using generalized linear modeling (GLM) and - in a ...
3
votes
0answers
147 views

Seasonal ARIMA Forecast

I'm studying ARIMA at the moment with application to seasonal data sets. R lets you forecast using selected models but I'm just wondering what formula is used to compute these forecasts. For example, ...
3
votes
0answers
2k views

How can i do time series forecasting with missing data

I am relatively new to time series forecasting, I have worked previously with continuous data at regular intervals successfully, Now I have a data set with missing values, for example look at the ...
3
votes
0answers
819 views

Forecastability and Coefficient of Variation

I'm trying to get a sense check here. When determining "forecastability" for sales data, I tend to use the CV. However, this is highly susceptible to seasonality and outliers. As such, I was wondering:...
3
votes
0answers
1k views

Parameter estimation for dynamic regression models with correlated noise ARMA errors

I'm reading the Dynamic Regression Models chapter ( https://www.otexts.org/fpp/9/1 ) in Professor Hyndman's book, and I couldn't understand how to fit the regression model when the error is modeled ...
3
votes
0answers
2k views

Stock closing price forecasting using ARIMA model in R

I have downloaded the daily stock Adjusted Close price of one stock from sep 2011 to till date. As per my study plan, I have plotted some basic plots to understand the daily stock Adjusted closing ...
3
votes
0answers
11k views

Removing seasonality from data

I have a dataset depicting weekly revenue over time for a computer company. The plot for the data looks like this: I decomposed the data into its additive components using the ...
3
votes
0answers
162 views

Verification of assumptions in TBATS model

I have a question about using BATS/TBATS models implemented in the forecast package for R. In De Liv­era, Hyndman & Snyder (2011) the models are used without any following analysis. Is it OK to ...
3
votes
0answers
159 views

Sales forecasting to account for regression

I have a very beginner question. I am attempting to forecast total 2014 unit sales of a large number of products. The data I have has 10 points for each individual product, which are the total unit ...
3
votes
0answers
224 views

Why is the Chow test with $R^2$ wrong?

The regular way to compute the F-value for a Chow forecast test is: $$F=\frac{(e_R'e_R-e_1'e_1)/g}{e_1'e_1/(n-k)}$$ My professor said something today about that a Chow forecast test using $R^2$ would ...
3
votes
0answers
138 views

Should I de-mean a predictor variable before a dummy interaction

Suppose I have the following time-series linear model where $\beta$ is misspecified: $Y(t+1) = \alpha + \beta X(t) + \sum_{i=1}^{10000}\gamma_i Z_i(T) + \varepsilon$ where all parameters are in $\...
3
votes
0answers
244 views

How do I forecast a timeseries of data using GARCH(1,1)?

I'm new to GARCH, but I've got daily data of TV Ratings. I've been trying to forecast this for future, and a quick background - the data is non-stationary, has high seasonality (weekly, monthly & ...
3
votes
0answers
1k views

R forecast package: How to combine fourier terms with another XREG matrix

I'm using R forecast package with a daily time series data, that has complex i.e. Multiple seasonality (weekly, Yearly, monthly). The fit/forecast process also needs to take into account certain day ...
3
votes
0answers
131 views

Prediction model problem

I am trying to design a model that can estimate the number of customers I will receive in every store every month using the number of customers I received every month in every store for the last five ...
3
votes
0answers
163 views

Modeling relative contribution of a variable

I am overthinking this for sure, but I am stumped. I have a historical data set of projects with hours of contribution by various positions. There are six types of projects. How can I model the ...
3
votes
0answers
66 views

Updating a set of estimated forecasts

Suppose I have some stochastic process $X_t$. At each time $t$, I receive an estimated probability distribution for $x_t$, followed by an observation $x_t$. After receiving a set of observations ${x_1,...
3
votes
1answer
526 views

Fitting time series with outliers

I have daily sales data for a department store for the past 850 days. I have indicators on the major holidays and the days leading up to the major holidays. The number of days before the holidays that ...
2
votes
0answers
10 views

Predict Sales as Counterfactual

Which modelling strategy (time frame, features, modelling technique) would you recommend to forecast 3-month sales for total customer base? At my company, we often analyse the effect of e.g. ...
2
votes
0answers
48 views

Resources for learning the time series stuff they don’t (or didn’t) teach you

I at one point, a long time ago, had two years of graduate econometrics focusing on time series, plus more on micro cross-section techniques. I haven’t made much use of the time-series stuff for a ...
2
votes
0answers
51 views

How to forecast low values in data more accurately than the higher values?

I have a scenario where I have to forecast small values in data more accurately than the higher values. I have data set as below ...
2
votes
1answer
49 views

95% prediction interval for an ARMA(2,2) model

What would the formula for a 95% prediction interval for an ARMA(2,2) model be? The specific model I am using is: an ARIMA(2,0,2) with non-zero mean, with the following parameter estimates: ...
2
votes
1answer
53 views

How does the R function Arima () calculate drift?

The Arima() function in the R forecast package contains an "include.drift" parameter. Could someone explain how this is calculated and how it is included in point forecasts? According to this post ...

1 2 3 4 5 21