INTRODUCTION:

In today’s world, there are a multitude of factors involved in running a business and working towards achieving an incredible growth path. With the processes becoming more complex in nature, there arise a need for leaner, faster and intelligence based decision making activities. Business processes and operational structures need to be efficient as well as effective monetarily. Businesses are processing and generating a huge amount of data with every passing second. Effectively using this data to drive better decision making, providing exemplary customer experience while trying to reduce the underlying costs is the biggest challenge for any organization¹.

The problem lies in creating systems and processes that can understand, process, translate and represent this huge amount of data for the betterment of the organization. The advancement in technology and the evolution of cutting edge technologies like Machine learning, BigData, IoT, Block chain, it has opened a door to drive business intelligence. The e-commerce and retail outlets are generating a massive amount of data with every transaction and machine learning provides an intelligent and scalable way to perform statistical analysis on this data using scientific methods and arrive at any conclusion².

Machine learning as a branch of Artificial intelligence is growing at a very rapid rate and business want to cash in any opportunity that they can leverage out of it. Its usage is now in but not limited to manufacturing, educational learning systems, healthcare, banking, agriculture, games etc.

Machine learning can be compared to a human learning. It learns from its experiences without necessary being told to perform a specific task. Machine learning employs the techniques of statistics and uses algorithms to devise a process where the machine tries to learn from its experiences without the intervention of any human being. The main aim of machine learning is to make computers learn automatically. Learning for a computer means findings patterns in data is provided with, process it, and conclude something meaningful³.

Working of Machine Learning Systems:

Machine learning works by providing the training data to a learning algorithm. The learning algorithm works on the data and using the inference mechanism, it generates a new set of rules. These set of rules are known as Machine learning model. Generation of a model depends on the kind of training dataset being used. Same learning algorithm can be used to generated different models depending on the input data provided to it⁴.

Firstly, the data is acquired from many sources like database, sensors, files and so on with large values, missing values or noisy data to obtain the outcome through time stamps and labels.

Once the data is prepared, The Training process is done by making the system learn with supervised or unsupervised methods like clustering, Classification etc. Hence an appropriate method is chosen to train the model. Since the method is selected, now we need to validate the method through machine learning by generating the results. If the results obtained are as desired, then the process is further send for the analysis.

Diagram 1: Stages of Machine Learning

In mathematical terms, this can be represented with a linear equation as:

y = f(x) + e

It can be described as creating a function which maps the input variable x to an output variable y with possibility of some error anomaly as e. Here, x can be understood as the input data. Function f can be referred to as the learning algorithm and the variable y is the resulting machine learning model getting generated. In order to minimize the value of e i.e. error value, the more data is passed to the learning algorithm, the more accurate is the resultant model.

Components of a Machine Learning System:

Diagram 2: Machine Learning systems

To the core of any machine learning system lies the machine learning algorithm and the input data, also known as training dataset. In this era of automation, building an enterprise level machine learning systems requires processes, build and deployment pipelines which can automate the whole process. Also, as any business employing the machine learning techniques would require to use more than one such systems. So, the need of processes which can iteratively perform such actions not just save the time, but also the cost required to again start from scratch⁷.

For businesses processing a large amount of data, a single machine would probably not be able to provide the computing power to execute the algorithm. A tool to scale up and down the systems in terms of infrastructure fulfills this requirement. The output of all such machine learning systems is to produce a machine learning model as an output which can be used to make predictive analysis⁸.

Machine Learning Algorithms:

There are different classifications of machine learning algorithms as shown below:

Diagram 3: Types of Machine Learning Algorithms^5,6

1. Supervised Learning: In this method of machine learning implementation, both input and output are provided to the algorithm. Feedback loops are also incorporated. For E.g., an image of a car could be provided as an input and output as a car, it will supervise or help the algorithm to identify the rules to classify any object as a car.

2. Un-supervised learning: In this type of machine learning method, no label is provided on the input data and the algorithm has to find its own ways to discover patterns in the input data.

3. Reinforcement learning: It works on a basic concept of rewards and punishments in which an agent interacts and operates within an environment and automatically determine the best possible action to maximize the performance. An agent is essentially a computer program.

Use of Machine Learning in Business:

Businesses today employs a number of sophisticated machine learning methods to aid in decision making, customer satisfaction, fraud detection and even the kind of music someone listens. These systems enable any organization to take informed decisions which are statically calculated and are way faster than the manual decision making processes^8,9,10,11. Businesses which employs the machine learning methods falls within but not limited to agriculture, medical and healthcare, security, fraud prevention, spam detection etc.^12,13,14.

Some of the major applications of machine learning across different businesses are^{,15,16,17,18,19}:

· Medical and Healthcare:

There are a number of applications of ML in medical and healthcare research, management, diagnosis etc. Cancer research is one such major area where medical researchers are trying to figure out the exact causes of the disease and also the ways to develop advanced cures. Machine learning methods are playing a major role in the research as for the development of any cure, humongous amount of data has to be processed and analyzed. IBM Watson Genomics is an example of such systems which are helping to detect these otherwise hard to diagnose diseases. These methods are mostly unsupervised learning as no previous patterns or results are known. ML methods are also used to predict chronic diseases, to find ways to reduce the hospital stay of any patient, to reduce the hospital readmissions etc.

· Agriculture:

In Indian context, agriculture is still mostly dependent on manual labor. Machine learning methods have gained a lot of momentum in agriculture fields recently which could be a game changer in the field. ML methods are being employed to create smart robots which can substitute a human being in the process of picking fruits, germination etc. Automating the irrigation systems is another important application which makes use of the data gathered from soil (moisture, mineral content etc), weather, crop cycles and many more factors to decide the amount of irrigation required for a better yield. This is very useful in recent times seeing the water scarcity problem across the world.

· Education:

ML systems are helping teachers in student’s assessment, gathering detailed insights about the topics where students are lacking, predicting student performance. It is also helping students to make use of customized learning materials according to the preferences and post analyzing the student expertise in a particular topic.

· Stock Trading Industry:

Companies using cutting edge technologies are using Machine learning models for stock market prediction and algorithmic trading. They heavily rely on the previous trade history, market growth and downfall patterns per stock, using sophisticated statistical techniques specially designed for stock data analysis. ML systems are particularly useful in making decision making very quickly as while doing the trades, even the milliseconds matter to book a good trade. It has provided the capability to automatically buy and sell stocks without the need of any human intervention.

· Oil and Gas:

Implementations in Oil and Gas are vast and providing cost cutting solutions across the industry. ML systems are being used to analyze the underground mineral deposits, finding the new energy sources, maintaining the oil and gas distribution channels, predicting the energy demands.

· Transportation:

Cab aggregators are using ML models in a number of activities including booking a cab, prediction of destination, location determination, optimized route based on the traffic data analysis and previous traffic patterns. This has resulted in more accurate pickup and drops.

· Retail:

The applications of ML systems in retail industry, both e-commerce and store based are humongous. Companies are using these systems to analyze consumer shopping patterns, suggestion engines, pricing strategies, spending pattern analysis etc. The data from products being purchased is used to build product development and marketing strategies.

· Credit Industry:

Credit industries being risk prone business have always used techniques to determine the credit worthiness before approving or denying the loans. This process used to be time taking due to the number of data points the companies had to analyze. ML systems are now being used in many places to determine the credit patterns of a consumer, spending patterns, current financial health and habits to process a loan application.

· Core Banking:

ML systems are helping banking institutions tackle with money laundering prevention, fraud detection, trade settlements, network security amongst many. These systems are helping banks cut time and money previously spent to perform these tedious tasks.

· Fashion Industry:

Within the fashion industry, ML systems are providing benefit to customers by providing them suggestions regarding the fashion to opt according to their body dimensions and fashion styles. In addition to this, fashion brands are getting useful insights from the data regarding customer choices, brands and styles that are more in demand, changes required to a particular product as per the fitting sizes.

· Customer Support Industry:

Chat-bots are the new common while connecting to any major customer service center. Instead of having a human at the end of chats, companies are engaging ML enabled chat-bot systems. These systems use the previous chat history as well as previous interaction pattern analysis to help and support the end customers.

· Security Industry:

Some of the applications of machine learning in security industry are Image recognition, voice recognition, surveillance. ML systems are proving to be very useful in large scale monitoring where the data to be processed are captured from various sources such as security cameras, voice recordings, drones and then fed into the ML systems to prevent any unwanted or illegal activity. As compared to humans, these systems are capable of processing the data very rapidly and providing more accurate results.

CONCLUSION:

Machine learning as a part of Artificial Intelligence can be compared to data mining to analyze and extract useful information from the available data. The business activities/organizations are dealing with huge amount of data on a daily basis. The growth of these business organizations depend on how qualitative they process this huge data for their growth. Using machine learning methods helps the organizations and industries to take better decisions quickly, define the growth strategy, perform predictive analysis, benefit monetarily in terms of cost reduction and perform operations similar to a human without actual human involvement. Machine learning is showing huge potential in driving future processes.

REFERENCES:

1. R. Cioffi, M. Travaglioni, G. Piscitelli, A. Petrillo and F. De Felice, "Artificial Intelligence and Machine Learning Applications in Smart Production: Progress, Trends, and Directions", Sustainability, vol. 12, no. 2, p. 492, 2020. Available: 10.3390/su12020492.

2. M. Attaran and P. Deb, "Machine Learning: The New 'Big Thing' for Competitive Advantage", International Journal of Knowledge Engineering and Data Mining, vol. 5, no. 1, p. 1, 2018. Available: 10.1504/ijkedm.2018.10015621.

3. Leo Dencelin X, Ramkumar T. Distributed Machine Learning Algorithms to classify Protein secondary structures for Drug Design – A Survey. Research J. Pharm. and Tech. 2017; 10(9): 3173-3180. doi: 10.5958/0974-360X.2017.00564.9

4. Das, K. and Beher, R.N. (2017) ‘A survey on machine learning: concept, algorithms and applications’, International Journal of Innovative Research in Computer and Communication Engineering, February, Vol. 5, No. 2, pp.1301–1309.

5. R. M. Balajee, K. Venkatesh. A Survey on Machine Learning Algorithms and finding the best out there for the considered seven Medical Data Sets Scenario. Research J. Pharm. and Tech. 2019; 12(6):3059-3062. doi: 10.5958/0974-360X.2019.00518.3

6. Megha Mishra, Vishnu Kumar Mishra, H.R. Sharma. An Impact of Machine Learning with Lexcio-Syntatics Features of Question Classification. Research J. Engineering and Tech. 3(4): Oct-Dec. 2012 page 327-331.

7. Carvalho, T.P.; Soares, F.A.A.M.N.; Vita, R.; da Francisco, P.R.; Basto, J.P.; Alcalá, S.G.S. A systematic literature review of machine learning methods applied to predictive maintenance. Comput. Ind. Eng. 2019, 1, 1–12.

8. N. Soni, E. Sharma, N. Singh and A. Kapoor, "Artificial Intelligence in Business: From Research and Innovation to Market Deployment", Procedia Computer Science, vol. 167, pp. 2200-2210, 2020. Available: 10.1016/j.procs.2020.03.272

9. Faggella, "How to Apply Machine Learning to Business Problems", Emerj, 2020. [Online]. Available: https://emerj.com/ai-executive-guides/how-to-apply-machine-learning-to-business-problems

10. C. Services, C. Tool, P. Courage, O. Offerings and D. Platform, "Real-World Benefits of Machine Learning in Healthcare", Health Catalyst, 2020. [Online]. Available: https://www.healthcatalyst.com/clinical-applications-of-machine-learning-in-healthcare.

11. Sindhu J, Renee Namratha. Impact of Artificial Intelligence in chosen Indian Commercial Bank –A Cost Benefit Analysis. Asian Journal of Management. 2019; 10(4):377-384. doi: 10.5958/2321-5763.2019.00057. X

12. F. Weber and R. Schütte, "A Domain-Oriented Analysis of the Impact of Machine Learning—The Case of Retailing", Big Data and Cognitive Computing, vol. 3, no. 1, p. 11, 2019. Available: 10.3390/bdcc3010011.

13. Meenakshi K, Safa M, Karthick T, Sivaranjani N. A Novel Study of Machine Learning Algorithms for Classifying Health Care Data. Research J. Pharm. and Tech. 2017; 10(5): 1429-1432. doi: 10.5958/0974-360X.2017.00253.0

14. L. M˘aru¸ster, "A machine learning approach to understand business processes", CIP-DATA LIBRARY TECHNISCHE UNIVERSITEIT EINDHOVEN, 2003, ISBN 90-386-1688-0 NUR 982.

15. Anitha A, Revathi SV, Jeevanantham S, Eliza Godwin E. Intrusion Detection System based on Artificial Intelligence. Int. J. Tech. 2017; 7(1): 20-24. doi: 10.5958/2231-3915.2017.00005.0

16. Machine Learning for Business", Medium, 2020. [Online]. Available: https://medium.com/mindsdb/machine-learning-for-business-c24cb198dbb5.

17. Wuest, T.; Weimer, D.; Irgens, C.; Thoben, K.D. Machine learning in manufacturing: Advantages, challenges, and applications. Prod. Manuf. Res. 2016, 4, 23–45.

18. S. Sasireka. Comparative analysis on video retrieval technique using machine learning. Research J. Science and Tech. 2019; 11(2):148-154. doi: 10.5958/2349-2988.2019.00022.6.

19. Rashmi R., Nirmal Raj VK. A Study on the Implementation and the Impact of Artificial Intelligence in Banking Processes. Asian Journal of Management. 2021; 12(1):47-54. doi: 10.5958/2321-5763.2021.00008.1.

Received on 30.05.2021 Modified on 24.06.2021

Research J. Science and Tech. 2021; 13(3):213-217.

DOI: 10.52711/2349-2988.2021.00033