Open AccessReducing Significances of Mesh Sensors Technologies through Dimensionality Reduction Algorithm
Author(s) -
Ruhul Amin
Publication year - 2020
Publication title -
engineering international
Language(s) - English
Resource type - Journals
ISSN - 2409-3629
DOI - 10.18034/ei.v8i2.556
Subject(s) - cuckoo search , computer science , dimensionality reduction , robustness (evolution) , cuckoo , data mining , reduction (mathematics) , machine learning , feature extraction , computation , algorithm , artificial intelligence , curse of dimensionality , feature selection , process (computing) , mathematics , zoology , biochemistry , chemistry , geometry , particle swarm optimization , biology , gene , operating system
In today's world, the breadth of real-time applications and networks is not limited to business and social activities. They are expanding as a field to provide improved and competitive settings for a variety of activities such as home, health, and commercial procedures. Data analytic method is used to maintain network accessibility as well as the robustness of expert services. It is necessary to clean up the data in order to reduce the computational complexity of extracting and pre-processing models. Because present approaches are sophisticated, they necessitate large computations. To this effect, the objective is to deploy a machine learning algorithm – “cuckoo search algorithm” for dimensionality reduction problems in data extraction for IoTs application. The cuckoo search-based feature extraction algorithm is a mutant algorithm that organizes itself depending on the unpredictable amount of input and generates a new and improved feature space. After the cuckoo search-based feature extraction is implemented, a few test benchmarks are provided to assess the performance of mutant cuckoo search algorithms. As a result of the low-dimensional data, classification accuracy is improved while complexity and expense are lowered.