Artificial Intelligence
Name
Institution Affiliation
Executive Summary
As the modern era of technology continues to witness significant advancements, several tools and machines have been created to best suit the human means of survival. One of the considerations made by the research is that the modern era has embraced technological progression where different machine culture has made a considerable impact on the broader scoop of applications which rages from making a proper and enhanced understanding of the text, making it easier for speech decoding and recognition of images (MARR, 2019). In this regard, several sectors, like healthcare and genomics, have also faced numerous advancements in their daily operation. This is due to the implication of artificial intelligence. Artificial intelligence is referred to as the wide-ranging use of computers to help in manipulating machines through simulating human intelligence and programming those machines to think like humans and mimic their actions. In order to fully implement these simulations in machines, the computers require several collections to function fully. Among these collections needed is the data that would help in analyzing, visualizing, and featuring in the engineering process. Although, in this simulation process, there several challenges that are encountered, among these challenges are machine learning and data acquisition processes (Thomas & Thomson, 2005).
Data collection in AI can be referred to as the process of acquiring and measuring information from different sources so that the information can be used in the creation of practical artificial intelligence and machine learning solutions. Machine learning (ML) is the study of computer algorithms that automatically improve through experience. Machine learning is an AI application that enables systems to learn and improve automatically from experience without being explicitly programmed Lawrence et al., 2016). It also offers a wide range of knowledge. Machine learning focuses on the development of computer programs that can access data and learn on their own. This research paper will entirely research on the existence of machine learning and different disciplines associated or that which are responsible for the data management for the purpose of machine simulation, and also aim to have a fair understanding of the data collection techniques that are suitable for such areas as semi-supervised learning, active learning coupled with transfer learning. Additionally, the survey will also include a comprehensive explanation that is towards seeking to identify the significant research challenges that exist.
To collect adequate and correct data for this research paper, it will use research tools that will be used in the survey. The research will require both software and hardware tools that would facilitate the survey. Among these tools needed are phone, laptop, and a reliable internet connection. This tool will help in having proper and smooth access to dominant scholarly sources that contain information such as ACM digital library, IEEE Xplore, and the dblb computer science bibliography. All these sources will be essential in surveying the expectation are that these materials contain key information on data acquisition, big data, and machine learning. National Library of Medicine (2019) clarifies that to deal with the growing changes in data structures and technology-driven research environment; the study recommends researching software and hardware tools to reimage research methodology programs to enable the researcher to develop appropriate competences to deal with the challenges of working with complex and large amounts of data associated analytics.
Table of Content
Executive Summary ………………………………………………………………………………………………. 2
Introduction and Review of the Literature ………………………………………………………………… 5
Problem statement and Research hypothesis …………………………………………………………….. 7
Methodology ………………………………………………………………………………………………………… 8
Results …………………………………………………………………………………………………………………. 9
Discussion ……………………………………………………………………………………………………………. 10
References ……………………………………………………………………………………………………………. 11
Introduction
Innovative solutions are needed in smart production systems to enhance quality and sustainability while reducing costs. In this context, I4.0 Key Enabling Technologies (e.g., the Internet of Things; advanced embedded systems, cloud-based computing; big data; cognitive systems; virtual and augmented reality) may help develop new industrial paradigms in the field of artificial intelligence technology (Artificial Intelligence and Machine Learning, 2019). In this regard, rapid advances in the field of artificial intelligence have profound implications for both the economy and society at large. The possibility for these innovations is that they can influence both the production of a wide range of products and services and their characteristics directly. But, as important as these effects are likely to be, artificial intelligence also has the potential to change the process of innovation itself, with consequences that may be equally profound, and that may come to dominate the direct effect over time. Machine Learning (ML) is used to conduct large data predictive analysis or pattern recognition (Zoubin, 2015). It also offers a series of tools to support alert and risk-management decisions to improve the overall operation of machines in taking up human functionality.
Review of the Literature
The techniques of machine learning have been leading in recent years as big data develops. As we had discussed earlier on, machine learning is a precise artificial intelligence (AI) area which aims to analyze the huge chunks of data and make it easier for the system to learn data without explicit programming support automatically. The learning algorithms attempt to reveal in multiple perspectives the fine grain patterns of unprecedented data and to create a precise prediction template like never before. For learning purposes, the machine learning algorithm is classified into four broad groups, including supervised education, semi-managed learning, uncontrolled learning, and enhancement education (Sankara, 2017). When new unseen data is provided as a computer-learning algorithm, it learns and forecasts what is coming automatically by taking advantage of past experience. Machine learning is constantly releasing its strength in a wide range of applications, including the Internet of Things (IoT), computer vision, natural language processing, speech processing, online recommendation, cybersecurity, psychology, predictive analysis, fraud detection, and so on.
The machine learning algorithm is capable of handling multifaceted data in a complex environment (Bonev, 2017). Given its pros, difficulties and challenges still exist. The machine learning algorithm also needs an additional method to predict a large number of new groups. Furthermore, the sophisticated analytical capabilities of the machine learning algorithm present new difficulties and challenges when maintaining privacy. The lack of properly recorded raw data makes the training process more difficult and contributes to imprecise tests. In addition, human experts are responsible for the efficient use of an algorithm for machine learning. Interpreting results is a major challenge in the algorithm of machine learning. The machine learning technique allows for hidden insights into large-scale data. However, the issue of high dimensionality, distributed computing, scalability, adaptability, and streaming data affects its operation. The duplication of data also has a major impact on the learning of the algorithm of machine learning. The classification of the multiple classes for the evolving large-scale data also induces complexity. The key problem of the machine learning algorithm is its susceptibility to error. In addition, there is a lack of variability in machine learning. The algorithm of machine learning provides accurate experience details to predict the future. In addition, large-scale data are needed to learn the different subjects that take a lot of time and resources.
Problem statement and research hypothesis
Even though data collection is a topic in machine learning, due to the increasing amount of training data, data management is now relevant, which has led to the merging of the two areas. Therefore, there is a need for awareness of the revolving research landscape for the communities and more efforts to incorporate the techniques. On the other hand, there are two main problems related to the Theory and Foundations of Artificial Intelligence (AI). One is to consider a problem-defining technique in AI. The other is to find a strategy for evaluating AI hypotheses. As of now, there are no solutions to these two problems. The former problem was ignored because researchers found it difficult to identify AI problems conventionally. The second issue has not been seriously discussed, with the result that potentially conflicting AI theories are usually non-comparable. If we are to suggest that our AI theories do have validity, it must be shown that we have evaluated them scientifically.
Despite the many application of machine learning, data collection is one of the stumbling blocks. It is an undisputed fact that a significant amount of time is spent on preparing data that encompasses, analyzing, visualizing, and feature engineering. Mosher and Freund (2018) state that data collection is a significant challenge because machine learning is applied in many applications, even though there is no enough training data. The earliest applications, such as machine translation or object detection, have vast amounts of training data at their disposal. However, recent applications suffer from little or no training data.
Methodology
The present research employed a case study design by utilizing both software and hardware tools to take on the study. Ali et al. (2017) argue that familiar sites and other needed tools were effectively used in taking part in the research. Case studies (both single and multiple) and empiric studies are the two predominant approaches for the study of AI, data management, and organization interactions. Most case studies are about successful applications; there is a shortage of failure cases, which may be even more important to analyze in relation to successes. Cases provide a practical perspective as well as the basis to establish scientific hypotheses that will ultimately be tested. The most commonly used method for the study of AI, data management, and organizations has to date been the review of a case only. Several published studies concentrate on a single application that has been popular. For example, the organizing impact of XCON in digital was defined by Cupallari (2020). The system improved the local efficiency of configuration work, and the system supports Digital’s product strategy directly, which is due to the use of XCON, which enhances its information processing capacity.
Results
This article provides an actor knowledge in the semi-structured format by analyzing the results of this model. This model addresses massive and operational open education data. These data are usually represented in unstructured format; according to large data architecture, they are distributed across multiple nodes. This proposal and the results that have been obtained show that we have a suitable model. This is because of the speed of these transactions resulting from the initial processing of massive data originally generated by actor interaction. In order to evaluate the existence of machine learning and different disciplines associated or that which are responsible for the data management for the purpose of machine simulation, the module research models were categorized into three categories;
Category 1 and 2: these are for knowledge that is dynamic or irrelevant. This is a tutorial to analyze the motivation of learners in Online Training.
Category 3: for relevant expertise representing the knowledge to be acquired in order to enhance the recommendation phase to give learning actors more autonomy in development stages. This model thus pre-processes the outcome of the relevant knowledge available to the ontological layer.
Discussion
Machines learning have established itself as one of the most popular technique in the area of Artificial Intelligence, and data management is an integral part of machines learning models as the performance of these models largely depend on user data. This research focused on the study of machine learning and how they relate to data management. The literature available on any subject is now wide, and complete coverage of all the documents published with respect to a particular topic can be challenging or even impossible. This document provides a systematic review of applications in various fields in AI techniques. The purpose of the research was aiming not only to provide a well outlined and comprehensive framework on the literature of the research of I but was also providing an explanation of ways towards seeking to identify the significant research challenges that still affect the field. As the research was mainly done using the software and hardware available, the documents mainly focused on taking deeper research. In saying these, several documents were visited. Among these documents was Google scholar, which had restricted access.
Bibliography
MARR, B. (2019). ARTIFICIAL INTELLIGENCE IN PRACTICE: how 50 successful companies used artificial … intelligence to solve problems. JOHN WILEY & Sons.
Ali, R., Lee, S., & Chung, T. (2017). Accurate multi-criteria decision-making methodology for recommending a machine learning algorithm. Expert Systems With Applications, 71(C), 257-278.
Artificial Intelligence and Machine Learning. (2019). Studies in Health Technology and Informatics, 261, 135.
Bonev, George. (2017). Machine Learning Algorithms for Automated Satellite Snow and Sea Ice Detection, The Graduate Center, All Dissertations, Theses, and Capstone Projects.
Cupallari, Andi. (2020). Applications of Machine Learning and Deep Learning in Macroeconomic and Financial Forecasting, The Graduate Center, All Dissertations, Theses, and Capstone Projects.
Lawrence, D., Palacios-Gonzãlez, C., & Harris, J. (2016). Artificial Intelligence. Cambridge Quarterly of Healthcare Ethics, 25(2), 250-261.
Mosher, J. H., & Freund, M. B. (2018). U.S. Patent Application No. 15/337,963.
National Library of Medicine. (2019). Artificial Intelligence and Machine Learning. Studies in Health Technology and Informatics, 261, 135.
Sankara Subbu, Ramesh. (2017). Brief Study of Classification Algorithms in Machine Learning, City College of New York, Master’s Theses.
Thomas, P., & Thomson, Gale. (2005). Artificial intelligence (Gale virtual reference library). Detroit, Mich.: Lucent Books.
Zoubin Ghahramani. (2015). Probabilistic machine learning and artificial intelligence. Nature, 521(7553), 452-459.