Computational Recovery of Sample Missings

2021 
Many problems in the real world include uneven label distributions where the proportion of samples of one type in the data is greatly different from the proportion of samples of other types. Due to the difficulty in the data collection, this data imbalance problem widely exists. For life science issues, this data sampling missing is difficult to alleviate by improving data-collection methods. Thus the problem of data imbalance poses a challenge to traditional machine learning methods. In response to this problem, this chapter proposes structure-aware and rebalancing learning. Through the analysis of internal structure of the data, the proposed method tries to rebalance the unbalanced data. On the association analysis and prediction tasks, we demonstrate the strucure-aware rebalancing method can efficiently improve the analysis of imbalanced data.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    50
    References
    0
    Citations
    NaN
    KQI
    []