Decoding the Future: A Comprehensive Review of Machine Learning Innovations and Applications

Abstract

In the current scenario of the 4th Industrial Revolution (4IR or Industry 4.0), the digital world is a full of data, such as Internet of Things (IoT) data, business data, mobile data, cyber security data, social media data, etc. To intelligently analyze these data and develop the corresponding smart and automated applications, the knowledge of artificial intelligence (AI), particularly, machine learning (ML) is the key. Supervised, unsupervised, semi-supervised and reinforcement learning are the different types of machine learning algorithms. In addition to the deep learning is part of a broader family of machine learning methods that can wisely analyze the data on a large scale. This study\'s primary contribution is its explanation of the fundamentals of numerous machine learning techniques and how they can be applied in a wide range of real-world application areas, including e-commerce, cyber security systems, smart cities, healthcare, and agriculture, among many others. The main use of machine learning is to show off its potential for generating consistently accurate estimations. This review paper\'s primary objective is to give an overview of machine learning and provide machine learning approaches.

Introduction

I. INTRODUCTION

We live in the age of data, where everything around us is connected to a data source, and everything in our lives is digitally recorded [1, 2].  For instance, the current electronic world has a wealth of various kinds of data, such as the Internet of Things (IoT) data, cyber security data [3], smart city data, business data, smart phone data, social media data, health data, COVID-19 data, and many more. ML usually provides systems with the ability to learn and enhance from experience automatically without being specifically programmed and is generally referred to as the most popular latest technologies in the fourth industrial revolution (4IR or Industry 4.0) [2, 3]. “Industry 4.0” [4] is typically the ongoing automation of conventional manufacturing and industrial practices, including exploratory data processing, using new smart technologies such as machine learning automation.

Machine learning is multi-disciplinary field in artificial intelligence, probability, statistics, information theory, philosophy, psychology, and neurobiology. Machine learning solves the real world problems by building a model that is good and useful approximation to the data. The study of Machine learning has grown from the efforts of exploring whether computers could learn to mimic the human brain, and a field of statistics to a broad discipline that has produced fundamental statistical computational theories of learning processes.

In 1946,  ENIAC, the first computer system was developed. The idea at that time was that mortal thinking and literacy could be rendered logically in such a machine. Alan Turing, in 1950, proposed a test to measure its performance. The Turing test is grounded on the idea that we can only determine if a machine can actually learn if we communicate with it and can not distinguish it from another human. Around 1952 Arthur Samuel (IBM) wrote the first game-playing program, for checkers, to achieve sufficient skill to challenge a world champion. In 1957 Frank Rosenblatt constructed the perception which connects a web of points where simple opinions are made that come together in the larger program to break more complex problems. In 1967, pattern recognition is developed when first program suitable to fete patterns were designed grounded on the type of algorithm called the nearest neighbor. In 1981, Gerold Dejong introduced explanation grounded literacy where previous knowledge of the world is handed by training exemplifications which makes the use of supervised literacy. In the early 90’s machine literacy came veritably popular again due to the crossroad of Computer Science and Statistics.

Advances continued in machine learning algorithm within the general areas of supervised and unsupervised learning [5]. In the present period, adaptive programming is in explored which makes use of machine literacy where programs are able of feting patterns, learning from experience, abstracting new information from data and optimizing the effectiveness and delicacy of its processing and affair.

Further interest is developed in machine literacy moment is because of growing volumes and kinds of available data, computational processing that's cheaper and more important, and affordable data storehouse. All of these effects mean it's possible to snappily and automatically produce models that can dissect bigger, more complex data and deliver briskly, more accurate results indeed on a veritably large scale. Machine literacy model produces high- value prognostications that can guide better opinions and smart conduct in real time without mortal intervention.

II. MACHINE LEARNING MODEL

The machine learning model's learning process is split into two parts:

1. Training
2. Testing.

Samples from the training data are used as input in the training phase, whereby the learner or learning algorithm learns the features and builds the learning model. During the testing phase, the learning model predicts the test or production data using the execution engine. The output of the learning model that provides the final forecast or classified data is called "tagged data."

III. MACHINE LEARNING TECHNIQUES

Machine Learning algorithms are mainly divided into four categories: Supervised learning, unsupervised learning, Semi-supervised learning, and Reinforcement learning [5].

A. Supervised learning

Supervised learning is typically the task of machine learning to learn a function that maps an input to an output based on sample input-output pairs [6]. Supervised learning is carried out when certain goals are identified to be accomplished from a certain set of inputs [3], i.e., a task-driven approach. Supervised Literacy is trained using labeled exemplifications, similar as an input where the asked affair is known. Supervised learning provides dataset consisting of both features and labels. For example, a piece of equipment could have training data points labeled either as F (failed) or as R (runs). Supervised learning is carried out when certain pretensions are linked to be fulfilled from a certain set of inputs.

Supervised learning is generally used in operations where literal data predicts likely unborn events. For illustration, it can anticipate when credit card deals are likely to be fraudulent or which insurance client is likely to file a claim. Other complicated examples includes recognition system as given a multicolor image of an object through a telescope, determine whether that object is a star,  or a galaxy, or given a list of movies a person has watched and their personal rating of the movie, recommend list of movies they would like.

Supervised Literacy tasks are divided into two orders as bracket and retrogression. In bracket, the marker is separate, while in retrogression, the marker is nonstop. For illustration, in astronomy, the task of determining whether an object is a star, a world, or a quasar is a bracket problem where the marker is from three distinct orders. On the other hand, in retrogression problem, the marker (age) is a nonstop volume, for illustration, chancing the age of an object grounded on compliances

B. Un-Supervised learning

Unsupervised learning analyzes unlabeled datasets without the need for human interference, i.e., a data-driven process [6]. This is extensively used for rooting generative features, relating meaningful trends and structures, groupings in results, and exploratory purposes. The most common unsupervised literacy tasks are clustering where a set of inputs is divided into groups, unlike in bracket, the groups aren't known before. Popular unsupervised ways include tone- organizing charts, nearest- neighbor mapping, k- means clustering and singular value corruption.

Unsupervised literacy used data that has no literal markers and the thing is to explore the data and find parallels between the objects. It's the fashion of discovering markers from the data itself. Unsupervised literacy works well on transactional data similar as identify parts of guests with analogous attributes who can also be treated also in marketing juggernauts. Or it can find the main attributes that separate client parts from each other.

Other unsupervised literacy problems are

1. Given detailed compliances of distant worlds, determine which features or combinations of features are most important in distinguishing between worlds.
2. Given an admixture of two sound sources for illustration, a person talking over some music, separate the two which is called the eyeless source separation problem.
3. Given a videotape, insulate a moving object and classify in relation to other moving objects which have been seen.

C. Semi-supervised Learning

Semi-supervised learning can be defined as a hybridization of the above-mentioned supervised and unsupervised methods, as it operates on both labeled and unlabeled data [6, 3]. In numerous practical literacy sphere similar as textbook processing, videotape indexing, bioinformatics, there's large force of unlabeled data but limited labeled data which can be precious to induce. So semi supervised literacy is used for the same operations as supervised literacy, as it operates on both labeled and unlabeled data. Therefore, it falls between learning “without supervision” and learning “with supervision”. There's an asked prediction problem but the model must learn the structures to organize the data as well as make prognostications. Semi-supervised literacy is useful when the cost associated with labeling is too high to allow for a completely labeled training process. This type of literacy can be used with styles similar as bracket, retrogression and prediction.

D. Reinforcement Learning

Reinforcement learning is a type of machine learning algorithm that enables software agents and machines to automatically evaluate the optimal behavior in a particular context or environment to improve its efficiency [8], i.e., an environment-driven approach. It's frequently used for robotics, gaming and navigation. It's the literacy fashion which interacts with a dynamic terrain in which it must perform a certain thing without a schoolteacher explicitly telling it whether it has come near to its thing. With underpinning literacy, the algorithm discovers through trial and error which conduct yield the topmost prices. So in the chess playing, underpinning literacy learns to play a game by playing against an opponent which performs trial and error conduct to win.

This type of literacy has three primary factors the learner, the terrain and conduct. The ideal is for the learner to choose conduct that maximize the anticipated price over a given quantum of time. The learner will reach the thing much briskly by following a good policy. So the thing in underpinning literacy is to learn the stylish policy.

IV. REVIEW OF LITERATURE

In the current age of the Fourth Industrial Revolution (4IR), machine literacy becomes popular in colorful operation areas, because of its literacy capabilities from the history and making intelligent opinions. In the following, we epitomize and bandy popular operation areas of machine literacy technology.

1. Cybersecurity and threat intelligence: Cybersecurity is one of the most essential areas of Industry 4.0. [4], which is generally the practice of guarding networks, systems, hardware, and data from digital attacks. Machine learning has come a pivotal cybersecurity technology that constantly learns by exploring data to identify patterns, better identify malware in encoded traffic, find insider pitfalls, predict where bad neighborhoods are online, keep people safe while browsing, or secure data in the cloud by uncovering suspicious activity. For instance, clustering ways can be used to identify cyber-anomalies, policy violations, etc. To detect various types of cyber-attacks or intrusions machine learning classification models by taking into account the impact of security features are useful [7].
2. Internet of things (IoT) and smart cities: Internet of Things (IoT) is another essential area of Industry 4.0. [4], which turns everyday objects into smart objects by allowing them to transmit data and automate tasks without the need for mortal commerce. IoT is, therefore, considered to be the big frontier that can enhance almost all activities in our lives, such as smart governance, smart home, education, communication, transportation, retail, agriculture, health care, business, and many more [9]. Smart city is one of IoT’s core fields of application, using technologies to enhance city services and residents’ living experiences [11,12]. As machine learning utilizes experience to recognize trends and create models that help predict future behavior and events, it has become a crucial technology for IoT applications [2].

For illustration, to prognosticate business in smart metropolises, parking vacuity prediction, estimate the total operation of energy of the citizens for a particular period, make environment- apprehensive and timely opinions for the people, etc. are some tasks that can be answered using machine literacy ways according to the current requirements of the people.

3. Traffic prediction and transportation: Transportation systems have come a pivotal element of every country’s profitable development. Nonetheless, several cities around the world are experiencing an excessive rise in traffic volume, resulting in serious issues such as delays, traffic congestion, higher fuel prices, increased CO2 pollution, accidents, emergencies, and a decline in modern society’s quality of life [13]. Therefore, an intelligent transportation system through prognosticating unborn business is important, which is a necessary part of a smart megacity. Accurate traffic prediction based on machine and deep learning modeling can help to minimize the issues [17, 18, 19]. For illustration, grounded on the trip history and trend of traveling through colorful routes, machine literacy can help transportation companies in prognosticating possible issues that may do on specific routes and recommending their guests to take a different path. Eventually, these literacy- grounded data- driven models help recover business inflow, increase the operation and effectiveness of sustainable modes of transportation, and limit real- world dislocation by modeling and imaging upcoming changes.

4. Healthcare and COVID-19 pandemic: Machine learning can help to solve diagnostic and prognostic problems in a variety of medical domains, such as disease prediction, medical knowledge extraction, detecting regularities in data, patient management, etc  [33, 32, 26]. Coronavirus disease (COVID-19) is an infectious disease caused by a newly discovered coronavirus, according to the World Health Organization (WHO) [15]. Recently, the learning techniques have become popular in the battle against COVID-19 [37]. For the COVID- 19 epidemic, the literacy ways are used to classify cases at high threat, their mortality rate, and other anomalies. It can also be used to better understand the virus’s origin, COVID-19 outbreak prediction, as well as for disease diagnosis and treatment [14, 22]. With the help of machine literacy, experimenters can read where and when, the COVID- 19 is likely to spread, and notify those regions to match the needed arrangements. Deep learning also provides exciting solutions to the problems of medical image processing and is seen as a crucial technique for potential applications, particularly for COVID-19 pandemic [10, 20, 21]. All things considered, machine learning and deep learning methods can aid in the fight against the COVID-19 virus, the pandemic, and wise clinical judgment in the healthcare industry.

5. E-commerce and product recommendations: One of the most popular and extensively utilized uses of machine learning is product suggestion, which is also one of the most noticeable aspects of practically every e-commerce website on the internet today. Businesses can use machine learning technology to help them analyze the past purchases of their customers and provide personalized product recommendations for their next purchase based on their behavior and preferences. For instance, e-commerce businesses can quickly position offers and product recommendations by looking at browsing patterns and the click-through rates of particular products. Using predictive modeling based on machine learning techniques, many online retailers, such as Amazon [34], can better manage inventory, prevent out-of-stock situations, and optimize logistics and warehousing. The ability to gather, assess, and use customer data to deliver a personalized shopping experience will be critical to the success of sales and marketing in the future. Additionally, businesses may retain current consumers while drawing in new ones by using machine learning techniques to develop content and packages that are customized to their needs.

6. NLP and sentiment analysis: Natural language processing (NLP) involves the reading and understanding of spoken or written language through the medium of a computer [30,2]. NLP thereby aids computers in a variety of tasks, such as reading text, listening to voice, interpreting it, assessing sentiment, and determining which elements are important enough to employ machine learning algorithms on. Virtual personal assistant, chat, speech recognition, language or machine translation, document description, etc. are examples of NLP-related tasks. Sentiment Analysis [31] (also referred to as opinion mining or emotion AI) is an NLP sub-field that seeks to identify and extract public mood and views within a given text through blogs, reviews, social media, forums, news, etc. Businesses and brands utilize sentiment analysis to gain insight into the social sentiment surrounding their brand, product, or service across social media platforms or the internet in general.

Sentiment analysis is generally regarded as a machine learning process that evaluates texts for polarity, encompassing "positive," "negative," or "neutral," and also more nuanced emotions such as very happy, happy, sad, very sad, angry, interested, or uninterested.

7. Image, speech and pattern recognition: Image recognition [36] is a widely recognized and commonly used application of machine learning in practical situations. This involves identifying an item depicted in a digital image. For example, distinguishing whether an x-ray shows signs of cancer, identifying characters, detecting faces in images, and providing tagging suggestions on social media platforms like Facebook are all typical instances of image recognition. Speech recognition [23] is also very popular that typically uses sound and linguistic models, e.g., Google Assistant, Cortana, Siri, Alexa, etc. [35]. These systems utilize machine learning methods. Pattern recognition [16] involves automatically identifying patterns and regularities in data, for example, analyzing images. Various machine learning methods, like classification, feature selection, clustering, and sequence labeling, are employed in this field.

8. Sustainable agriculture: Agriculture is essential to the survival of all human activities [29]. Sustainable agriculture practices help to improve agricultural productivity while also reducing negative impacts on the environment [40, 25, 27]. The sustainable agriculture supply chains are knowledge-intensive and based on information, skills, technologies, etc. [18], where knowledge transfer encourages farmers to enhance their decisions to adopt sustainable agriculture practices utilizing the increasing amount of data captured by emerging technologies, e.g., the Internet of Things (IoT), mobile technologies and devices, etc. [40,38,39]. A variety of stages of sustainable agriculture can benefit from the application of machine learning. These include the pre-production phase, which involves forecasting crop yield, soil characteristics, irrigation needs, etc.; the production phase, which involves forecasting weather, disease and weed detection, managing soil nutrients, managing livestock, etc.; the processing phase, which involves estimating demand and planning production; and the distribution phase, which involves inventory management and consumer analysis, etc.

9. User behavior analytics and context-aware smartphone applications: Context-awareness is a system’s ability to capture knowledge about its surroundings at any moment and modify behaviors accordingly [28, 93]. Software and hardware are used in context-aware computing to automatically gather and interpret data for immediate answers. The mobile app development environment has been changed greatly with the power of AI, particularly, machine learning techniques through their learning capabilities from contextual data [2]. Thus, the developers of mobile apps can rely on machine learning to create smart apps that can understand human behavior, support, and entertain users [24, 27]. Machine learning techniques are applicable to build a variety of personalized data-driven context-aware systems, such as smart mobile recommendation, smart interruption management, context-aware smart searching, and decision-making that intelligently assist end mobile phone users in a pervasive computing environment. An intelligent phone call application, for instance, can be developed using context-aware association rules. By considering data from time series, clustering techniques are helpful in capturing the variety of behavioral activities that users engage in. Classification techniques can be used to forecast future events in a variety of scenarios.

Conclusion

We have performed a thorough overview of machine learning algorithms for applications and intelligent data analysis in this paper. In line with our objective, we have covered in brief the ways in which diverse machine learning techniques can be applied to solve a range of real-world problems. The performance of the learning algorithms and the data are both necessary for a machine learning model to be successful. Before the system can support intelligent decision-making, the complex learning algorithms must first be trained using the real-world data and knowledge relevant to the intended applications that have been gathered. Additionally, in order to demonstrate the machine learning techniques\' relevance to a range of real-world problems, we talked about a few well-liked application areas. We have now covered the main points of the difficulties encountered, as well as possible avenues for future research and directions in the field. As a result, the challenges that have been identified present exciting opportunities for field research and need to be addressed with practical solutions across a range of application areas. All things considered, we think that our research on machine learning-based solutions points in a promising direction and can serve as a technical reference manual for future studies and applications for professionals in academia and business.

References

[1] Cao L. Data science: a comprehensive overview. ACM Comput Surv (CSUR). 2017;50(3):43. [2] Sarker IH, Hoque MM, MdK Uddin, Tawfeeq A. Mobile data science and intelligent apps: concepts, ai-based modeling and research directions. Mob Netw Appl, pages 1–19, 2020. [3] Sarker IH, Kayes ASM, Badsha S, Alqahtani H, Watters P, Ng A. Cybersecurity data science: an overview from machine learning perspective. J Big Data. 2020;7(1):1–29. [4] ?lusarczyk B. Industry 4.0: Are we ready? Polish J Manag Stud. 17, 2018. [5] Mohammed M, Khan MB, Bashier Mohammed BE. Machine learning: algorithms and applications. CRC Press; 2016. [6] Han J, Pei J, Kamber M. Data mining: concepts and techniques. Amsterdam: Elsevier; 2011. [7] Kaelbling LP, Littman ML, Moore AW. Reinforcement learning: a survey. J Artif Intell Res. 1996;4:237–85. [8] Sarker IH, Abushark YB, Alsolami F, Khan A. Intrudtree: a machine learning based cyber security intrusion detection model. Symmetry. 2020;12(5):754. [9] Mahdavinejad MS, Rezvan M, Barekatain M, Adibi P, Barnaghi P, Sheth AP. Machine learning for internet of things data analysis: a survey. Digit Commun Netw. 2018;4(3):161–75. [10] Alakus TB, Turkoglu I. Comparison of deep learning approaches to predict covid-19 infection. Chaos Solit Fract. 2020;140: [11] Zanella A, Bui N, Castellani A, Vangelista L, Zorzi M. Internet of things for smart cities. IEEE Internet Things J. 2014;1(1):22–32. [12] Zheng Y, Rajasegarar S, Leckie C. Parking availability prediction for sensor-enabled car parks in smart cities. In: Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP), 2015 IEEE Tenth International Conference on. IEEE, 2015; pages 1–6. [13] Guerrero-Ibáñez J, Zeadally S, Contreras-Castillo J. Sensor technologies for intelligent transportation systems. Sensors. 2018;18(4):1212. [14] Ardabili SF, Mosavi A, Ghamisi P, Ferdinand F, Varkonyi-Koczy AR, Reuter U, Rabczuk T, Atkinson PM. Covid-19 outbreak prediction with machine learning. Algorithms. 2020;13(10):249. [15] World health organization: WHO. http://www.who.int/. [16] Anzai Y. Pattern recognition and machine learning. Elsevier; 2012 [17] Boukerche A, Wang J. Machine learning-based traffic prediction models for intelligent transportation systems. Comput Netw. 2020;181 [18] Essien A, Petrounias I, Sampaio P, Sampaio S. Improving urban traffic speed prediction using data source fusion and deep learning. In: 2019 IEEE International Conference on Big Data and Smart Computing (BigComp). IEEE. 2019: 1–8. [19] Essien A, Petrounias I, Sampaio P, Sampaio S. A deep-learning model for urban traffic flow prediction with traffic events mined from twitter. In: World Wide Web, 2020: 1–24 [20] Yujin O, Park S, Ye JC. Deep learning covid-19 features on cxr using limited training data sets. IEEE Trans Med Imaging. 2020;39(8):2688–700. [21] Shorten C, Khoshgoftaar TM, Furht B. Deep learning applications for covid-19. J Big Data. 2021;8(1):1–54. [22] Jamshidi M, Lalbakhsh A, Talla J, Peroutka Z, Hadjilooei F, Lalbakhsh P, Jamshidi M, La Spada L, Mirmozafari M, Dehghani M, et al. Artificial intelligence and covid-19: deep learning approaches for diagnosis and treatment. IEEE Access. 2020;8:109581–95. [23] Chiu C-C, Sainath TN, Wu Y, Prabhavalkar R, Nguyen P, Chen Z, Kannan A, Weiss RJ, Rao K, Gonina E, et al. State-of-the-art speech recognition with sequence-to-sequence models. In: 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2018 pages 4774–4778. IEEE. [24] Zulkernain S, Madiraju P, Ahamed S, Stamm K. A mobile intelligent interruption management system. J UCS. 2010;16(15):2060–80. [25] Cobuloglu H, Büyüktahtak?n IE. A stochastic multi-criteria decision analysis for sustainable biomass crop selection. Expert Syst Appl. 2015;42(15–16):6065–74. [26] Gökhan S, Nevin Y. Data analysis in health and big data: a machine learning medical diagnosis model based on patients’ complaints. Commun Stat Theory Methods. 2019;1–10 [27] Sharma R, Kamble SS, Gunasekaran A, Kumar V, Kumar A. A systematic literature review on machine learning applications for sustainable agriculture supply chain performance. Comput Oper Res. 2020;119: [28] Dey AK. Understanding and using context. Person Ubiquit Comput. 2001;5(1):4–7. [29] Sarker IH, Salah K. Appspred: predicting context-aware smartphone apps using random forest learning. Internet Things. 2019;8: [30] Otter DW, Medina JR , Kalita JK. A survey of the usages of deep learning for natural language processing. IEEE Trans Neural Netw Learn Syst. 2020. [31] Ravi K, Ravi V. A survey on opinion mining and sentiment analysis: tasks, approaches and applications. Knowl Syst. 2015;89:14–46. [32] Nilashi M, Ibrahim OB, Ahmadi H, Shahmoradi L. An analytical method for diseases prediction using machine learning techniques. Comput Chem Eng. 2017;106:212–23. [33] Fatima M, Pasha M, et al. Survey of machine learning algorithms for disease diagnostic. J Intell Learn Syst Appl. 2017;9(01):1. [34] Marchand A, Marx P. Automated product recommendations with preference-based explanations. J Retail. 2020;96(3):328–43. [35] López G, Quesada L, Guerrero LA. Alexa vs. siri vs. cortana vs. google assistant: a comparison of speech-based natural user interfaces. In: International Conference on Applied Human Factors and Ergonomics, Springer. 2017; 241–250. [36] Fujiyoshi H, Hirakawa T, Yamashita T. Deep learning-based image recognition for autonomous driving. IATSS Res. 2019;43(4):244–52. [37] Lalmuanawma S, Hussain J, Chhakchhuak L. Applications of machine learning and artificial intelligence for covid-19 (sars-cov-2) pandemic: a review. Chaos Sol Fract. 2020:110059 . [38] Kamble SS, Gunasekaran A, Gawankar SA. Sustainable industry 4.0 framework: a systematic literature review identifying the current trends and future perspectives. Process Saf Environ Protect. 2018;117:408–25. [39] Kamble SS, Gunasekaran A, Gawankar SA. Achieving sustainable performance in a data-driven agriculture supply chain: a review for research and applications. Int J Prod Econ. 2020;219:179–94. [40] Adnan N, Nordin Shahrina Md, Rahman I, Noor A. The effects of knowledge transfer on farmers decision making toward sustainable agriculture practices. World J Sci Technol Sustain Dev. 2018.

Copyright

Copyright © 2024 Bhavika C. Donga, Piyush D. Pitroda, Dr. Hasmukh B. Domadiya, Dr. Dipti H. Domadiya. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.