Analysis of Bayesian optimization algorithms for big data classif i cation based on Map Reduce frameworkChitrakant Banchhor * and N. Srinivasu IntroductionT h e data size is increased because of the technological developments in the new data computing fi eld. Big data refers to the large dataset...
Most learn-ing algorithms are based on statistics, i.e. they relyon a large amount of data. However, humans canlearn tasks and coherences from very few (or evenone) examples.Of course, if we strive to augment statisticallearning with other methods, we have to be awarethat the information...
The third module verif i es our G1Nos algorithms bycomparing it with two of the most used SMOTE-like oversampling al-gorithms evaluating the diagnostic ability of a binary classif i er system(Multi Layer Perceptron).The experimental results show that our oversampling algorithm workbetter than ...
Ahmad SR, Bakar AA, Yaakub MR (2015) Metaheuristic algorithms for feature selection in sentiment analysis. In: Science and information conference (SAI), 2015. IEEE, pp 222–226 Ali F, Kwak D, Khan P, Islam SR, Kim KH, Kwak KS (2017) Fuzzy ontology-based sentiment analysis of transport...
Thus, an interesting research direction for sys- tem deployment in practical scenarios is to develop algorithms that make the conver- gence fast, stable, and adaptive by properly controlling the regularization parameter V and the queues evolution step sizes and η. We may also adapt the resource ...
In this section, we experimentally compare the presented algorithms among themselves. The reason for this is that, to the best of our knowledge, there is no other published clustering algorithm in the literature for data streams implemented on top of the MapReduce model (which provides scalability...
The increasing volume and velocity of the continuously generated data (data stream) challenge machine learning algorithms, which must evolve to fit real-world problems. The data stream clustering algorithms face issues such as the rapidly increasing volu
KS and SPXY are typical Euclidean distance-based algorithms for spectral sample selection. SSK is a commonly used K-nearest neighbor-based IS algorithm. RS, used to verify the optimal samples, is more effective than randomly selected samples with the same size as LARIS. These four methods are ...
In the process of developing algorithms, the functional form of the algorithm is determined, the function of the parameters is derived from a set of input and output pairs and these data have an impact on the scope of the trained data set and the model performance; besides, the non-...
The former is a class of algorithms that attempt to maximize the community structure recovery while the latter considers the process of generating a network that exhibits community structure with high fidelity. In this section, we briefly explain recent work on these algorithms. 2.1. Discriminative ...