|
@@ -363,7 +363,7 @@ as your buckets:
|
|
|
"buckets_path": "the_sum",
|
|
|
"window" : 30,
|
|
|
"model" : "simple",
|
|
|
- "predict" 10
|
|
|
+ "predict" : 10
|
|
|
}
|
|
|
}
|
|
|
--------------------------------------------------
|
|
@@ -445,4 +445,4 @@ minimization is linear to the size of the window being processed: excessively la
|
|
|
Finally, minimization fits the model to the last `n` values, where `n = window`. This generally produces
|
|
|
better forecasts into the future, since the parameters are tuned around the end of the series. It can, however, generate
|
|
|
poorer fitting moving averages at the beginning of the series.
|
|
|
-======
|
|
|
+======
|