Forecasting Models:
A Brief Explanation to Its Ability
There are many modelling algorithms nowadays, several of them could be categorized to forecasting models. Simple Moving Average (SMA), Auto-regressive Integrated Moving Average (ARIMA), Random Forest, and Long Short-Term Memory (LSTM) are some of the forecasting models. Different with other prediction models, forecasting models are used when we have a dataset with a time dimensional feature.
But, what can this model do? Let’s take a look at several implementations of forecasting models, below!
Weather Forecasting
Well, let’s start with a classic implementation. When we watch morning news on the TVs, they often show how the weather on the day would be; whether it be rainy, cloudy, or sunshine. How could they say the weather condition even before it happened? Yup, they use the forecasting model. Weather is a natural event and has a seasonal condition. It means, by analyzing past trends, we could predict how the weather would be this day, tomorrow, or even a month later.
Stock Prediction
Forecasting models could be used in the economic field as well. A big company would have a dashboard showing their stock condition. And not only that, a predictive simulation of the stock could also be displayed here. A stock data would have a timestamp feature updated at a determined duration, usually in minutes. With that amount of data, a forecasting model could be built to predict the stock condition in the next time. Based on the prediction, stakeholders could make a better decision when managing the company stock.
Climate Change Projection
Recently, there is a scientific publication about the global warming issue. The Intergovernmental Panel on Climate Change (IPCC) just released a report containing several graphs of global climate projections. I am not going to discuss the report substance, but, well you guessed it: the projections are built with a forecasting model. In the report, IPCC shows the projections of climate conditions from 1990 to…
…no, not today or decades, but 2300!
Of course there would be an unpredictable event that could mess-up the prediction, that is why it is called ‘prediction’. But, the interesting point I want to mention here is that with a proper dataset, we could even predict more than a hundred years of the future condition.
And more…
There are many more implementations of forecasting models in the real world. Forecasting could be used to predict from a short-term to long-term of time, as well as the stakeholder need and availability of the data.
