Skip to main content

Table 2 Uncertainty mitigation strategies

From: Uncertainty in big data analytics: survey, opportunities, and challenges

Artificial intelligence Uncertainty Mitigation
Machine learning Incomplete training samples
Inconsistent classification
Learning from low veracity and noisy data
Active learning [65, 66], Deep learning [15, 63], Fuzzy sets [67], Feature selection [9, 60, 61]
Learning from unlabeled data Active learning [65, 66]
Scalability Distributed learning [12, 63]
Deep learning [56]
Natural language processing Keyword search Fuzzy, Bayesian [68, 70, 71]
Ambiguity of words in POS ICA [73], LIBLINEAR and MNB algorithm [68]
Classification (simplifying language assumption) ICA [73], Open issue [68]
Computational intelligence Low veracity, complex and noisy data Fuzzy logic, EA [76, 79, 80, 82]
High volume, variety Swarm intelligence, EA [78, 81, 82], Fuzzy-logic based matching algorithm, EA [81, 82]