"PCA for 101010101 series prediction"
Guru
Dear All,
I have a process which predicts the next Boolean value given a Boolean series.
The processes first applies windowing.
Then a sliding window validation is ran.
Inside the sliding window validation PCA is applied.
After this integer {1, 0} values are converted Boolean.
And then a J48 learner is applied.
Before apply model {1, 0} values are converted yet again converted to Boolean.
This constantly converting between Boolean to integer makes the processes really slow!
Is there a way to overcome this problem?
Can we apply PCA to a Boolean data set?
Best regards,
Wessel
< portSpacing端口= " sink_through 1”间隔= " 0 " / >
<运营商激活= " true " class = " apply_model“compatibility="5.1.006" expanded="true" height="76" name="Apply Model" width="90" x="180" y="30">
I have a process which predicts the next Boolean value given a Boolean series.
The processes first applies windowing.
Then a sliding window validation is ran.
Inside the sliding window validation PCA is applied.
After this integer {1, 0} values are converted Boolean.
And then a J48 learner is applied.
Before apply model {1, 0} values are converted yet again converted to Boolean.
This constantly converting between Boolean to integer makes the processes really slow!
Is there a way to overcome this problem?
Can we apply PCA to a Boolean data set?
Best regards,
Wessel
< portSpacing端口= " sink_through 1”间隔= " 0 " / >
<运营商激活= " true " class = " apply_model“compatibility="5.1.006" expanded="true" height="76" name="Apply Model" width="90" x="180" y="30">
0
Answers
actually you simply can leave the integer as integer. The J48 will convert them to bins itself. Or do I overlook something?
Greetings,
Sebastian