MicroRNAs (miRNAs) take part a significant role in cancer development and also act as a vital feature in several other diseases. Previously a standard classifier method like SVM classifier exploited for selecting promising miRNAs encompass differential expression in benign and malignant tissue samples. Consequently, the nondominated sets of capable miRNAs are combined into a single most promising miRNA subset. On the other hand, the most important drawback of such learning techniques is slow learning time. With the aim of overcoming these problems of conventional learning techniques, in this work Extreme Learning Machine classifier is formulated for deciding promising MiRNAs because it only needs modification of one parameter. The performance has been demonstrated on four real-life miRNA expression datasets for ELM and the identified miRNA markers are reported. The experimental results demonstrate that the proposed ELM method outperforms the standard methods.