Incomplete Time Series Prediction Using Max-Margin Classification of Data with Absent Features

Figure 1

Two distinct processes of incomplete time series prediction.

(a) Traditional process

(b) Our process