HOW TO PREPARE YOUR DATA
First and foremost, the first column is for the predicted variable. In the example, it is the "output".
Please ensure that the first row is the earliest observation, from the first period, and the last row is the newest observation. For example, in the table below, the first period when the observer started recording, there was no output, input#1 was 2 units and input#2 was at 50%. On the last period, again there was no output, input#1 was 0 units and input#2 was at 100%.
Output |
Input#1 |
Input#2 |
0 |
2 |
50 |
0 |
1 |
80 |
0 |
5 |
50 |
0 |
2 |
80 |
0 |
3 |
20 |
10 |
3 |
100 |
15 |
2 |
100 |
20 |
3 |
100 |
0 |
0 |
100 |
With the Flexible App #2, you can make a prediction for the "Output" variable in the table below. The prediction will be whether the impact will be "Good", "Very_good" or "Excellent." Note that the result cannot be "Okay", because there is no past information about that in the dataset.
Please notice the underscore "_" instead of a space " " in values, e.g., "Very_good" instead of "Very good".
Output |
Input#1 |
Input#2 |
Good |
2 |
50 |
Good |
1 |
80 |
Good |
5 |
50 |
Good |
2 |
80 |
Very_good |
-2 |
20 |
Excellent |
-2 |
100 |
Excellent |
2 |
100 |
Excellent |
-2 |
100 |
Good |
-5 |
100 |