Accuracy & Prediction of your forecast: Throw the dart to find out.

There is no forecast that is 100 percent correct. The difference between the forecast and the actuals is known as the forecast error, which consists of both systematic and random error.

• Random error is a error which differs between the observations without consistency or it cannot be attributed to any variable.
• Systematic error is a consistent or proportional difference between the observed and true values of something.

?Let's look at the Dart experiment to further grasp what these errors mean. Different darts are thrown many times by various participants in this experiment. With each throw, the goal is to aim to strike the target's center. The bullseye is the point at which darts are most accurate.

In the above diagram,

1. There is no accuracy nor precision, if the darts are neither close to the bullseye nor to one another.
2. There is precision but not accuracy if every dart lands distant from the bulls-eye but very closely together.
3. Because the average of the darts lands in the bullseye, there is mathematical precision if the darts are all thrown at roughly the same angle from and spread equally around the target. This info is accurate but not precisely defined.
4. Both accuracy and precision are present if the darts land close to the target and closely spaced apart.

Precision, which is how accurately the same measurement can be made again under similar conditions, is the major outcome of random error. Systematic error, on the other hand, has an impact on a measurement's accuracy, or how closely the observed value resembles the true value.

?In theory, random errors in a time series can be described as an unknowable random variable. It is extremely difficult to eliminate random errors in a systematic manner.

?Typically, a systematic error will result in more accurate data than a random error. Data that is exceedingly accurate but wrong in particular circumstances could be preferred to data that is accurate but imprecise. Highly accurate measurements that are significantly wrong are frequently the result of systemic errors, which can then be fixed by improving the forecasting and model training procedures.