Nwokonkwo Obi Chukwuemeka,John-Otumu Adetokunbo MacGregor,Nnadi Leonard Chukwualuka,Ogene Ferguson
Issue:
Volume 7, Issue 1, June 2022
Pages:
1-7
Received:
20 December 2021
Accepted:
8 January 2022
Published:
14 January 2022
DOI:
10.11648/j.mlr.20220701.11
Downloads:
Views:
Abstract: Automated machine learning (AutoML) models is one of several machine learning algorithms that can be used to automate the solution of real-world problems. It automates the selection, composition, and parameterization processes of the machine learning models in particular. Machine learning could be more user-friendly when it is automated, and it often produces faster, and more accurate results than hand-coded machine learning methods. For more than ten years, AutoML for supervised learning has been the main focus of research under the discipline of artificial intelligence, and significant progress has been made theeafter; consider the usefulness of AutoML methods in the most popular machine learning toolkits, as well as the AutoML mechanisms in large scale platforms such as Microsoft Azure. This paper provides a methodical analysis of the AutoML workflow as well as the state-of-the-art effort in dealing with the challenges involving Combined Algorithm Selection and Hyperparameter Optimization by gathering information about AutoML from several published articles from different online repositories in order to delve more into the methods used in different domains and the level of accuracy obtained. Findings revealed that the next generation of machine learning and artificial intelligence research is focused on automating the other phases of the whole end-to-end machine learning pipeline, from data comprehension to model deployment. With significantly better deep learning algorithms and big datasets, AutoML is predicted to be able to handle most of the data cleaning process in the future. AutoML will evolve into a highly human-competitive system that will change the way we think about data research.Abstract: Automated machine learning (AutoML) models is one of several machine learning algorithms that can be used to automate the solution of real-world problems. It automates the selection, composition, and parameterization processes of the machine learning models in particular. Machine learning could be more user-friendly when it is automated, and it oft...Show More
Abstract: The research has designed a system that has done a morphological analysis of noun phrase and compound verb. Also, the system designed will translate a whole sentence indicating which words are noun and verb in it. Clustering was an unsupervised technique which was used to translate from English to Igbo language. In order to obtain our desired motives, object oriented analysis and design methodology were used. The system has been developed to make Igbo populaces to communicate well with most spoken English country along the global and strengthen the Igbo’s pole position in terms of research excellence. Furthermore, it will remove barriers to international trade that will keep Igbo small and medium companies from obtaining their complete economic standard by making ways into markets in other continents beyond our own. These goals lead us to develop a machine learning algorithm for translation of English into Igbo language. Machine learning algorithm for translation of English to Igbo language is the missing puzzle that will bring businesses to the people’s doorsteps. Besides, people that refused to acquire Igbo language are denying themselves pleasure of direct and unfiltered communication with others and thereby imprisoned themselves with the thrown of language.Abstract: The research has designed a system that has done a morphological analysis of noun phrase and compound verb. Also, the system designed will translate a whole sentence indicating which words are noun and verb in it. Clustering was an unsupervised technique which was used to translate from English to Igbo language. In order to obtain our desired motiv...Show More