Comments (5)
If you can represent the 'initial plan' as a value for each step in the historical data and in the forecast horizon, then you can add it as an exogenous variable to the forecasting process. The model will learn to take it into account when generating the forecast.
Otherwise, you would have to generate the forecast without the additional information and then combine it with the initial plan. But this combination would be external to the modelling process.
Hope this help
from skforecast.
Hi @gehal4
I don't know if I understand your question correctly, but it seems to be a problem related to the weight that each observation has during the training process.
If this is the case, this user guide may help you:
https://skforecast.org/latest/user_guides/weighted-time-series-forecasting
Please let me know if this is what you are looking for 😄
from skforecast.
Hi @gehal4
When you say "initial plan", do you mean an initial estimated forecast for each of the steps in the future?
from skforecast.
Hi @JavierEscobarOrtiz, @JoaquinAmatRodrigo,
"Initial plan" could also be the demand in retail, but also a plan of the future which is expecting a certain behavior due to the usage planned (referring as an example to a machine which might perform due to high/low usage as planned).
So as this "plan/demand" will not reflect 100% the realty due to many reasons (weather, workload changes etc.. etc.. different field different reasons), by taking into consideration both the history and this "plan/demand", how to produce a more realistic forecast ?
Hope my question is a bit clearer :)
from skforecast.
I see your point, and I think it's a interesting idea, I'll probably try both of them separately, for the second one an average between the plan and the generated forecast, and also both at the same time to test which one is giving the best accuracy.
Thanks for your comments @JoaquinAmatRodrigo and @JavierEscobarOrtiz !
from skforecast.
Related Issues (20)
- A single model multivariate forecaster HOT 3
- Issue saving ForecasterSarimax object HOT 4
- Custom predictors are inefficient for window features HOT 1
- grid_search_sarimax takes a very long time to run HOT 1
- Feature request: Add ability to skip steps in backtesting HOT 6
- Bad error handling when Index is neither RangeIndex nor DateIndex HOT 3
- Why do i have faster runtime when using more frequent refits HOT 2
- syntax error !! HOT 1
- Feature request: Custom predictors for multivariate forecasting
- Series-specifc exogenous vars for ForecasterAutoregMultiSeries HOT 2
- add method to forecasters to return the input data that is passed to the model to make predictions HOT 1
- allow extraction of features from y
- Feature request: Allow the training set to be passed to custom error metrics HOT 2
- XGboost using GPU
- MissingValuesExogWarning despite no NaN HOT 2
- Fitted Values HOT 2
- Using skforecast with panel data HOT 5
- Disable progress bar HOT 1
- Time Series Differentiation not working as expected HOT 3
- ForecasterSarimax - AttributeError: 'PeriodIndex' object has no attribute 'step' HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from skforecast.