Questions tagged [forecasting]
Prediction of the future events. It is a special case of [prediction], in the context of [time-series].
1,005
questions with no upvoted or accepted answers
2
votes
1answer
99 views
Should I avoid mixed ARMA models?
I have hourly demand data for taxi rides that spans several years into the past. I want to use it in order to forecast future demand (for the next day). Robert Nau warns against the usage of a mixed ...
2
votes
0answers
216 views
Estimating prediction interval of ARMA process using R forecast function
the theme is forecasting with ARMA models.
I'm trying to understand how the R forecast function works if applied to an ...
2
votes
1answer
434 views
How to predict weekly or monthly sales from daily time series model?
I've been given daily data and I've trained a SARIMAX time series model in Python so that I can predict daily data if given daily input.
However, I need to forecast on a monthly or weekly level, ...
2
votes
0answers
45 views
Is stationarity a requirement when using neural networks for time series forecasting?
I'm getting conflicting information on whether stationarity is a requirement when using neural networks for time series forecasting:
In this lecture, the speaker says it isn't a requirement.
In this ...
2
votes
0answers
26 views
ARIMA predictors - clarification
I'm working on multivariate time series (still), and would like some clarification. I was reading this site: Duke Forecasting and I came across this statement:
"We see that the most significant ...
2
votes
0answers
38 views
Forecasting using MA(2) model when past 5 observations are known
So given an MA(2) model : Xt = Wt + Theta1 * Wt-1 + Theta2 * Wt-2
Where Wt is white noise. (Normally distributed) and Theta1 and theta2 were available.
Say if X96,X97,...X100 of the series were given ...
2
votes
1answer
182 views
Forecasting with AR(1) and pseudo out-of-sample using R
I'm trying to do Pseudo out-of-sample forecasting using R. And, I also have the following initial data (gdp)
...
2
votes
0answers
210 views
The Efficient Market Hypothesis and forecastability?
According to Wikipedia:
The efficient-market hypothesis (EMH) is a theory in financial economics that states that asset prices fully reflect all available information. A direct implication is that ...
2
votes
0answers
75 views
auto.arima picks seasonal model for non seasonal series
I have a monthly time series data of 3 years whose acf and pacf plots confirms absence of seasonality. But auto.arima picks a seasonal model by seasonal difference first and then seasonal AR component ...
2
votes
0answers
149 views
Obtaining from scratch the volatility in GARCH model using R?
I'm trying to obtain the same vector of volatility by myself $\sqrt{h_{t|t-1}}$ of a Garch Model, that I obtained "automatically" using the function "ugarchfit" from the package "rugarch".
So after ...
2
votes
1answer
207 views
Flat forecast of trended time series data in r
I have a monthly time series of online visits for last 3 years starting from Jan 2016 to Dec 2018 and need to forecast for 2019. The data clearly has an upward trend although no seasonal lags ...
2
votes
0answers
80 views
Can the sum of several time-series be a white noise process, when the individual time series are not?
Intuitively, I think that it is possible for a sum of time series to be white noise, when the individual time series are not. Reason I am asking, is because I want to know if it's useful to ...
2
votes
0answers
195 views
Combinef function in R HTS pakage
First I would like to thank R people & package developpers for making available such a collection of great tools.
My question is about forecasts reconciliation, as I find huge differences ...
2
votes
1answer
37 views
How can a person predicted best playing 11 in a match between two teams?
This website allows people to bet on cricket and football matches. They ask people to select 11 players and there are point system, so at the end whoever ends with more points gets lots of money.
...
2
votes
0answers
39 views
Time Series Prediction Model for Home Prices
I am building a time series model to predict the zillow home prices for march 2019.I have data for each zip code from the year 1993 - 2018 and i have prices for every month.I was trying to use ARIIMA ...
2
votes
0answers
46 views
Using maximum of forecasted values to forecast maximum
I am using an algortihm to generate a daily sales Forecast and have concluded that the Forecast is, for pratical purposes, of good enough quality ("low" wMAPE).
In general, and without further ...
2
votes
1answer
289 views
Adding noise to time series data to increase training data
I am dealing with a weekly time series forecasting problem and I am currently investigating the use of an LSTM to make a multi-step forecast for a univariate time series. I actually have a ...
2
votes
0answers
83 views
How can I normalize truncated variables for a neural network?
Generally, I normalize variables using standard normal variates or (x-xmin)/(xmax-xmin)
But this only works well for variables that are not truncated, for example ...
2
votes
2answers
160 views
Rolling Window Forecasting with ARIMAX while supplying actual values
I am comparing different exogenous variables in how good they support the forecast of the monthly seasonal adjusted unemployment rate. All my data is monthly (2006-01-01 until 2018-09-01) and ...
2
votes
0answers
160 views
Is the forecasting model I am using overfitting and what is the best place to end training?
I am working on a forecasting model for natural gas consumption. I have many exogenous variables and when I train the data with the nnetar model(using R and the forecast packagae), one can specify the ...
2
votes
0answers
147 views
Convergence of predictions of an autoregressive model
I have performed a simple autogregressive model with lag 2 on a time series data. After obtaining the coefficients, I have computed the predictions. Since the lag is 2 in model, the first prediction $\...
2
votes
0answers
65 views
Forecasting autoregressive model. What's the best linear predictor?
Obviously if $X_t = \phi X_{t-1} + Z_t$, then the best linear predictor of $X_t$ given $X_{t-1}$ is $X_t = \phi X_{t-1}$.
But if $\phi$ is unknown, one may attempt to substitute $\phi$ by a Yule-...
2
votes
0answers
126 views
Interpreting forecast predictions of log transformed data
Using the forecast function in R, I make a 1-step prediction for a log-transformed data set Y, ( Y = log(X) ). This prediction gives me a mean and a 95% prediction interval.
How valid is this ...
2
votes
0answers
46 views
Are there any other models besides ARMA models that require stationarity?
Every now and then I come across a discussion of forecasting methods that mentions the topic of stationary time series vaguely without specifying that it is a question mainly in the context of ARMA ...
2
votes
0answers
93 views
How to interpret model confidence set in R
I want to compare 8 different forecast models to each other. Since I dont want to run into the $\alpha$-Inflation of multiple testing I heard about the model confidence set form Hansen.
I did this ...
2
votes
1answer
98 views
How can I split my time-series data into Train/Validation and Test set to apply Rolling Window
I am dealing with the LASSO regression in (pure AR-regressions) context. I have a lot of observations (around 4000). Therefore I would try to use the train/validation/test method. The idea was to ...
2
votes
0answers
47 views
Computing a corrective regression forecasting factor
I am working on forecasting problem using a regression model like gradient boosting to predict the number of weekly sold shoes.
I am using the historical data only from last year to predict the sales ...
2
votes
0answers
44 views
Evaluating which forecasting method works better? Statistical or Business Forecast
Somewhat new to the forecasting area. I am trying to evaluate whether the statistical forecasts are better than manually generated forecasts in one of our used cases. I have 1000s of customers who ...
2
votes
0answers
35 views
Testing time-series forecasts against actual observations
I'm conducting an event study on annual executive salaries.
I have a sample of 52 companies which have been given a cartel fine during year 6 (Event year).
For each company, I have a time series of ...
2
votes
1answer
254 views
ARIMA(1,1,1) Model - Forecast
How does one
write the mathematical equation for the ARIMA(1,1,1) model with the estimated coefficients below and
use the ARIMA(1,1,1) model and time series points below to produce a forecast value ...
2
votes
0answers
178 views
How can a combination of Random Forest and Linear Regression improve a time series forecast?
I attended a presentation by some consultants for retail demand forecasting who showed that for one of their clients, they were able to improve their demand forecasting by replacing a traditional time ...
2
votes
1answer
32 views
Establishing the minimum required training set size, when cross validating time series data
I want to evaluate and compare how well various models perform with regards to modelling time series data (the data in question is daily revenue). It seems that cross validation error might be a ...
2
votes
0answers
235 views
How to forecast hierarchical time series with external unique external regressors for each base time series?
I have hierarchical time series with 70 base time series, forming 4 level of hierarchies. I am using forecast() function in R from the package forecast.
The ...
2
votes
0answers
38 views
Out-of-sample forecasts: Why does model with log-transformed variables perform so much better?
I am developing a model to forecast the number of students enrolled in roughly 65 primary schools in a large city. Relevant predictors include the number of appropriately aged children living in the ...
2
votes
0answers
196 views
Implementaiton of Continuous Ranked Probability Score (CRPS) when Observation is a Distribution
The most general form of the Continuous Ranked Probability Score (CRPS) is defined as,
$\int_{\mathbb{R}} \big( \hat{F}^e(x) - F^0(x)\big)^2dx,$
for some true distribution, $F^0$, and empirical ...
2
votes
0answers
96 views
Explaining stationarity in a visual way
I am new to forecasting and want to try and explain to my peers in a visual and simple way how you know if a time series is stationary or not.
In the forecasting books I have read, the advice is ...
2
votes
1answer
54 views
Effect of strong auto-correlation on forecasting?
Suppose a wise-sense stationary univariate time series has relatively strong auto-correlation of lag-length of 1, say, around -0.7
Then how would it affect the forecast?
Conversely, if a ...
2
votes
0answers
77 views
Assign less weight to most current observations in forecasting
in forecasting, typically, we assign a heavier weight to the most current observations. However, I am finding many cases where a "blip" in last month's sales leads to a very pessimistic view about the ...
2
votes
0answers
159 views
Why does stl() decomposition require integer frequency?
I need to decompose and forecast weekly series with around 10 years of data. In this data leap years play an important role so I need the have non-integer frequency, frequency = (365.25/7)
By ...
2
votes
0answers
148 views
Are there any rules of thumb for the number of hidden layer neurons in a RNN or LSTM for time series prediction?
Say that I have a univariate time series X(t) that I want to forecast using RNN/LSTM.
I have 2 years of weekly sales data that is seasonal. How many hidden layers and neurons in each layer do I need ...
2
votes
0answers
405 views
Accuracy measures in training/test split of time series
I'm using Forecast Principles and Practice 2 to study time series and a doubt came in mind while I was trying to do exercise 7 of chapter 3.
How sensitive are the accuracy measures to the training/...
2
votes
0answers
62 views
Forecasting costs with forecast interval using past performance
I'm trying to adopt a model for project cost forecasting in agile.
Consider the following table of previous costs per sprint, along with story points completed:
...
2
votes
0answers
181 views
Tree model does't go well on trend
I am using sales time series data 2011 onwards, to make predictions for upto 2 years. Other than date and holiday related features, i created moving averages, y/y ratios and lags. I also extracted ...
2
votes
1answer
100 views
ARIMA forecasting using exogenous variables with their own forecast intervals
Suppose
model <- Arima(y , xreg=cbind(x1, x2), order=(p,d,q))
If I am forecasting $x_1$ and $x_2$, then for forecasting $y$:
1) If I use expected forecasts ...
2
votes
0answers
278 views
Is there any interpretation of parameters in Holt Winters method?
I am doing forecast on time series on R and I use exponential smoothing method Holt Winters.
Does a value of $\alpha$ close to $0$ or $1$ "mean" something particular about the series?
Same question ...
2
votes
0answers
86 views
What machine learning techniques to use to predict for multiple seperate sequences of time-series data?
I am having difficulty structuring my data and finding a machine learning technique to predict my outcome.
My data: I have a number of users with observations of a number of factors each year, each ...
2
votes
0answers
266 views
How much do the parameters in the Holt-Winters model matter?
When fitting a Holt-Winters model, I usually take the approach of retrospectively "predicting" some known historical values for the series, and optimising the coefficients for the parameters by ...
2
votes
1answer
350 views
A time series logit model with lagged dependent variable
I have a panel dataset for stocks. My goal is to model and predict if the stock will close positive (1) tomorrow based on today's close (1/0) and other macroeconomic and firm-specific variables.So I ...
2
votes
0answers
490 views
Alternatives to Holt-Winters models when the seasonality pattern has changed
I am forecasting a series of daily volumes in terms of units processed for a particular time period (the period around Christmas).
Historically, I have used a Holt-Winters model, with the minor ...
2
votes
0answers
266 views
Building the covariance matrix for hts prediction intervals
In my previous question: Using information about covariance between ARIMA models in forecasting I was interested in the more general case of how to use the covariance matrix in prediction intervals ...
