Tutorial 02 - Time-sequenced Data Preprocessing on RapidMiner

RapidMiner has strong coverage on manufacturing industries where we need to deal with multi-modal sensor data toward objectives like downtime, throughput, and yields. If we were to take the dataset as is to train a model directly, the result would have been very poor. But, if we were to visualize these sensor readings as waveforms, we might be alluded to deriving some attributes that are more interesting, for example, peak information, sub-window aggregations, and approximating coefficients.