Questions tagged [forecasting]
Prediction of the future events. It is a special case of [prediction], in the context of [time-series].
1,437
questions with no upvoted or accepted answers
2
votes
1
answer
2k
views
Restricting a set of predictions to a range of values of non-negative numbers
I am not even sure how to even phrase this question so if anyone could help that would be great.
I am analyzing facebook activity and I wish to predict a particular activity (comments, for instance). ...
2
votes
1
answer
192
views
Predicting water levels based on rainfall stats
I am curious if R or any other open source code can deal with forecasting changes in water elevation based on a predicted/forecasted value of rain.
I have a ton of data that shows water elevations (...
1
vote
0
answers
22
views
ETS Confidence Intervals in R are several orders of magnitude larger than the time series itself?
I am forecasting a time series with confidence intervals using the ets model in R. Here is the time series:
Running the following R code:
...
1
vote
0
answers
11
views
Is there a correction for samples from a (linear) Prophet model when trained on an inhomogenous Poisson point process?
Facebook's Prophet is a popular modelling choice for time series forecasting in production due to many steps being automated (and thus convenient). This can sometimes lead to over-reliance on it when ...
1
vote
0
answers
29
views
Multilayer Perceptron vs. Recurrent Neural Network for Time Series Forecasting: Utilizing Multiple Lagged Values
I am currently analyzing daily sales data for a product sold across multiple stores using a Multilayer Perceptron (MLP) model. For simplicity, let's assume it consists of a single layer, structured as ...
1
vote
0
answers
37
views
Decomposition time series
I am studying a daily dataset from a game which contains informations about the peak of players from 2013-2023. This is my first try applying decomposition to see how time series components behaves. ...
1
vote
0
answers
35
views
Valid forms of exploratory data analysis for time series that don't assume stationarity?
Lets say we are given a time series sample and want to try to create a model to forecast future values of said time series
When trying to build a model to forecast time series data, many statistics ...
1
vote
0
answers
39
views
Model fit / Forecast accuracy / Predictors / Explanatory power predictors (panel data)
I have the following data structure: 100 individuals (forecasters) predicted the likelihood of the outcomes of 50 events (binary outcomes, 1 or 0). For each event, each forecaster made two different ...
1
vote
0
answers
25
views
How to forecast changepoints from Gas Concentration Data?
So I'm trying to predict when gas concentrations change from sensor conductivity readings over a day. The gases randomly change concentrations around every 80-120 seconds and are kept constant between ...
1
vote
1
answer
65
views
Predicting quantiy sold using Time series data
I am struggling with a time series dataset comprising 12 features, including quantity sold and weather data, totaling approximately 1800 values, where data is recorded on a daily basis. My goal has ...
1
vote
0
answers
16
views
How to deal with different orders of integration between explained and explaining variables?
Is there a standard, or at least a valid, regression approach if you are trying to regress a dependent variable with a unit root against a set of stationary independent variables? I know I could ...
1
vote
0
answers
50
views
Multi-step ahead forecasting using ML models with horizon feature
All academic literature I could find on this topic distinguishes between recursive forecasting (forecasting y_{t+1}, then using it as an input to forecast y_{t+2} etc) or direct forecasting using a ...
1
vote
0
answers
34
views
regression model with arma errors: forecasting the residuals
Suppose I estimate the following model:
$$
y_t = \beta_0 + \beta_1 x_t + \eta_t
$$
where $\eta_t$ is an AR(1) model, say. I can do that with forecast::Arima() as ...
1
vote
0
answers
30
views
Enhancing Short-Term Sensitivity in Daily-Level Forecasting with Facebook Prophet
I have a daily time series dataset, and I'm using Facebook Prophet for daily-level predictions. Frequently, my actual data experiences sudden spikes due to external events. What I want to achieve is ...
1
vote
0
answers
46
views
The Hot Dog Vendor: Daily or Weekly Probability When Forecasting?
As a production planner, a variant of the following example has been bothering me for a long time; hopefully someone here can help me to view the problem in the right manner. Let's consider a ...
1
vote
0
answers
52
views
How does weather forecasting work in production settings?
The way I understand time series models is that you usually have to .fit() them every time you want to make a forecast. This is because there might be trends that were not seen before and so you can't ...
1
vote
0
answers
19
views
Singular Spectrum Analysis Grouping / Forecasting
I had a specific question about forecasting with Singular Spectrum Analysis and the impact that the grouping task has on this.
I understand that the grouping phase is important for interpreting the ...
1
vote
0
answers
26
views
Time series iterated multi-step forecasting autocorrelation issue (propagation of error)
I am trying to model some (quite ugly) time series data with lagged values of the outcome variable using a non-parametric machine learning model as I do not know how to correctly specify the model. ...
1
vote
0
answers
58
views
Backshift Notation for a Seasonal ARIMA model with constant
When fitting and forecasting using a seasonal ARIMA model with a total differencing of 1 (non-seasonal + seasonal), a constant term is necessary to capture the trend of the original data prior to ...
1
vote
0
answers
59
views
Is there a closed form for multi-step ARIMA/ARMA density forecasts conditioned on initial values?/alternatives to this?
I am attempting to create a benchmark for probabilistic forecasting of time series to test other models against and figured that a linear ARIMA/ARMA model would be a good starting point.
I thought ...
1
vote
0
answers
15
views
MSE for ARIMA and UR in linear models
I'm trying to assess the predictive power of some time series models (LASSO, ARIMA, UR) on time series data, but I have a problem.
I am conducting simulations with $N$ data points, $P$ potential ...
1
vote
0
answers
313
views
How to Reduce False Negatives? XGBoost Model with Imbalanced Dataset
I'm participating in a competition of binary classification of disease and I'm using an XGBoost model to classify my data. However, I am experiencing a higher number of false negatives than I would ...
1
vote
1
answer
28
views
Machine learning forecasting / asking for input parameter
I've been working on a weekly web traffic forecast based on time series. The primary business driver is marketing investment (spending)
I have historical spending data but not future spending data. In ...
1
vote
0
answers
70
views
How to model user behavior with a time series of discrete locations in continuous time?
I have a dataset that contains the sequential locations a user has visited, with start and end times. A small sample:
...
1
vote
1
answer
865
views
Time series split (expanding window) vs k-fold cross validation in time series forecasting
I do time series forecasting (q-o-q GDP) using ml models.
For hyper-parameter tunning I use grid search with cross-validation. Cross-validation is specifically Time series split (using expanding ...
1
vote
0
answers
61
views
How to forecast from longitudinal (panel) data
How is longitudinal forecasting usually done?
I know a bit about longitudinal (panel) data analyses, but I have not yet forecasted from longitudinal data.
I tend to think that building several time ...
1
vote
0
answers
44
views
Algorithm choose for: event based time series prediction
I have a dataset which is a electronic bus charger station, data include: voltage current and power consumption. There is no time rule in the use of charging piles, but I want to predict the power ...
1
vote
0
answers
24
views
What are the most common fault prediction algorithms?
I have to predict a fault (automotive related) as much in advance as possible.
Right now I have found a solution that is somewhat satisfactory (a good number of true positives and a low number of ...
1
vote
2
answers
179
views
Changepoint detection and forecasting
I am seeking to develop a three-staged algorithm that should be able to detect change in
Mean
Variance
Both Mean and Variance simultaneously if they exist.
I would be glad if this program can also ...
1
vote
0
answers
43
views
How to improve the accuracy of time series forecast using stlf()
I am new to working with time series and have tried several methods on my data including SARIMAX, Croston and forecastHybrid. The most accurcate result I've gotten so far is with stlf(), based on ...
1
vote
0
answers
29
views
Seasonal differencing applied to exogenous variable (xreg)? Forecast package R by Hyndman
Im currently working on specifying a seasonal ARIMA model with an exogenous variable. I'm using the forecast package developed by Hyndman for this. I have specified the following:
...
1
vote
0
answers
136
views
Interpreting lagged exogenous variables in ARMAX and regression with ARMA errors
There is an interesting post about the connection of lagged exogenous variables and the autoregressive time series model: Forecasting - Lags vs. AR terms for Exogenous Variables
Consequently, by using ...
1
vote
0
answers
58
views
Need help in random forest time series forecast
I am a beginner in time series forecasting using ML, and I am encountering a strange phenomenon. I have air quality data, in which I have information of various pollutants. The goal is to predict AIR ...
1
vote
1
answer
108
views
auto_arima straight line prediction python
I know this has been asked a lot but I have checked everything and still don't understand. To start, I have a dataset of global temperatures averaged over years. There is a trend in the series and I ...
1
vote
0
answers
96
views
ARIMA, VAR and State Space Model (SSM) forecasting comparison
I am trying to compare the asset price forecasting abilities of SSMs with ARIMA and VAR models. To keep it brief, this is the plan that I am following:
Collect multivariate data
Perform ADF ...
1
vote
0
answers
62
views
What discrete vector timeseries modeling (e.g. autoregression) methods support "continuity" requirements?
This question is motivated by the need to do vector-valued discrete time series forecasting with some guarantees of "continuity" (or rather, a discretized analog of continuity expressed in ...
1
vote
0
answers
51
views
RMSE of Training data is lower compared to test dataset
First of all, this is not a case of Overfitting.
The task is to forecast Temp using univariate Single Step forecasting.
I have trained the LSTM model with on jena climate dataset(Dataset https://...
1
vote
0
answers
18
views
Predicting future price from sales data
I want to be able to predict what a product's price would be based on its current sales listings and its sales history.
The historical sales data would be a list of (date, price, quantity)
The current ...
1
vote
0
answers
82
views
Calculating weights for a weighted least squares regression on a differenced time series
I am looking for guidance on whether I am approaching my problem correctly.
I have an annual time series { $x_{1}$, $x_{2}$, ..., $x_{t-1}$, $x_{t}$ }, where each observation is the estimated median ...
1
vote
0
answers
344
views
Measuring cannibalisation of category sales from the introducing of more brands/products
A retail company has traditional brick-and-mortar stores of varying sizes and assortments and a webshop with a considerably larger assortment.
I am interested in a method for measuring how the ...
1
vote
0
answers
79
views
Build time series models for 50k customers?
I have time series data for 50k customers.I want to forecast at a customer level. But the problem is it is not possible to train a ARIMA/ARIMAX model on each customer. Can I train a general time ...
1
vote
0
answers
15
views
Problem formulation of future timeframe prediction based on current time
I have a problem where I want to predict "when is the next action happening" based on the time.
Example problem: Imagine you have a dataset of transactions per user, your goal is to predict ...
1
vote
0
answers
18
views
auto.hd.type = "cv" in mlp() function in nnfor package
In running mlp() function of the nnfor package, you can allow the model to choose the number of hidden nodes through cross validation.
...
1
vote
0
answers
34
views
How to generate quantile forecasts from first differences?
Let's say I have a time series and I am taking the first differences and training a model to output the predicted 95% quantiles of these first differences at future time horizons. If this was just a ...
1
vote
0
answers
48
views
Can I use a paired t-test to assess whether there are statistically significant differences between the backtesting results of two forecasting models?
I developed a new forecasting model with the aim of replacing an older model. I conducted 12 backtests and need to predict only one quarter ahead. Overall, my new model performs better than the older ...
1
vote
0
answers
35
views
What is the suitable Graph Machine Learning model for dynamic graph forecasting with changing nodes and edges features
I have a graph where each node carries a feature value(s). The edges in the graph are weighted and each edge carries a single value (weight). The weight of the edge is some value calculated using the ...
1
vote
0
answers
242
views
Demand classification
Most of the industries use a following approach to classify the demand pattern.
Smooth demand (ADI < 1.32 and CV² < 0.49).
Intermittent demand (ADI >= 1.32 and CV² < 0.49).
Erratic demand (...
1
vote
0
answers
27
views
Which metric for neural network should I try for time series data with sudden peaks?
I am doing time series forecasting with neural network (feedforward for now, but I will test also RNNs) and my problem is that, even though the network learned general patterns, it doesn't forecast ...
1
vote
0
answers
20
views
Partial autocorrelation significant at regular lag distance
I am trying to forecast inflation with a simple AR model.
I took the natural logarithm of the CPI and subtracted the 12th lag, thus obtaining a measure of inflation.
The PACF is significant at the ...
1
vote
0
answers
192
views
Which models to use when forecasting time series data that shows exponential decay?
I'm working through "Forecasting: Principles and Practice (3rd edition)" by Rob J Hyndman and George Athanasopoulos to better understand times series forecasting in an R environment. This ...