A prediction model of integrated learning based on FCM and SAGA

Existing methods for big data analysis and prediction face difficulties due to the high dimensionality and disorderliness of big data, which cause noise problems. Moreover, the proportion of base learners in integrated learning often leads to redundancy. To address these issues, this paper proposes an integrated learning prediction model based on FCM-SAGA, which dynamically adjusts the learners by using the SAGA algorithm to achieve the optimal solution of the selector. For feature selection, this paper adopts FCM clustering to reduce the dimensionality and ensure that data clusters with similar variation curves can be aggregated together. Finally, the dataset selects real incremental one-year data to verify the advantages of the proposed algorithm over other mainstream algorithms in terms of accuracy.