Applications of Machine Learning

The use of AI can be seen anywhere. It is very possible that you are unknowingly using it in your daily routine. A very common application of AI is ML, in which different software operating systems work by perception (just like the human brain). Some examples of ML are used in our daily life without necessarily knowing that they operate via ML. ML has vast applications in real life. There are many day-to-day scenarios

ML applications

FIGURE 13.9 ML applications.

where ML is applied without the knowledge of the layman user. Determining the relevance of data is the job of ML; the machine is programmed in such a way that it is able to identify the relevance of data by using a variety of sources such as the web address, the number of users frequently browsing particular web pages, how often such web pages appear in similar searches, and the frequency of occurrence of the search string in the entire web page. All this information combines to help the machine learning algorithms (MLAs) understand how relevant a particular web page is for a given search (Intro Books #450, Machine Learning) (Figure 13.9).

Virtual Assistants

A virtual assistant is like a person who provides various services from a nearby location. Assistants used in smartphones, such as Bixby, Siri, Google Assistant, and Amazon Alexa, are a few of the most popular helpful personal assistants. They assist us by providing information when operated using one’s registered voice. All we have to do is switch them on and give commands or questions such as, “What are my appointments today?” "How is the weather in Paris?” or similar questions. To get you the results, your assistant looks through the data, memorizes your searched problems and their solutions, or send an instruction to other apps on your devices to gather information. We could even ask the assistant to perform some work such as, “Set a reminder for tomorrow’s meeting” or “Remember to visit Granny.” ML is an essential part of these types of virtual assistants because they gather and filter out data based on your previous activities with them. These virtual assistants are connected to a number of different items. A few of them are: [1]


Traffic situations: We usually have GPS access in our devices and we use it very commonly for our convenience. When we use the GPS in our devices we allow companies to track our current speed and location, and all this data is collected at a main server. The information is then processed to make an overview of the latest traffic conditions. These help to control traffic and prevent traffic jams, but the main problem is that there are not many cars or vehicles that have access to a GPS by themselves. ML in these situations is beneficial to find the places where overcrowd- ing is seen based on daily experiences.

Transport networks: Transport networks include taxis, rideshares, and public transit, for example. When we look for a taxi or a cab, now an app calls one and even provides the fare. When these services are shared, the question is how they could lessen the deflection? The solution is ML. The engineering lead at Uber, Jeff Schneider, disclosed that they use ML to explain price rush hours by determining customers demand. Now ML plays a vital role in our lives.

Videos Supervision

Let’s think of a person operating numerous surveillance cameras! Such a hectic job. That is why the prospect of upgrading PCs or other devices to perform this job makes sense. The video inspection strategies these days are built with AI that helps to find problems before they occur. They check out negative behavior like standing at a place continuously for a long time, tripping over, or taking nap w'hile sitting etc. The AI detects the activity and sends a report about that particular person, which helps to avoid any kind of issues. When such activities are submitted and found to be true, it helps them to further improve the monitoring. These all work together with the help of ML as a back support internally without coming into consideration of the operators.

Social Media Activities

  • Targeted advertisements: Making the newsfeed better by properly targeting advertisements, these sources are making use of ML either for purposes of their own or to help users. This is one of the well-known real-world examples of ML that you face on daily basis, without knowing that you see these fabulous things because of ML.
  • People You Should Know: The concept of ML is to learn from previous experiences. One the best-known things about Facebook, which does a great deal of ML, is that it constantly monitors the activities of people, such as where do we go, who do we meet, what are our field of interests, where do we work, or who do we like to hangout with. After analyzing all these things, Facebook shows us a bunch of people who also have similar tastes as we do.
  • Face Recognition: Suppose you are uploading a pic of your family and you find that Facebook very quickly identifies the people in the picture. Facebook locates the views and style in the pic uploaded, sees their uniqueness, and then it matches with each person in our friend list. This process seems to us as very easy but the truth is that all these tasks are very complicated and only made easy by the application of ML.

Spam E-mail and Filtering of Malware

A lot of techniques are used by people to identify spam e-mails. Resolving this issue means identifying spam files, so they are continuously upgrading their systems, which are equipped with ML. When junk filtering is completed the process must be repeated, as the junk file creators adopt the latest techniques to produce these mails. Multilayer perceptron and C4.5 decision tree induction are two of the junk-detecting methods that are equipped w'ith ML.

Supporting Customers Online

There are numerous online sites these days that offer the choice of communicating with a representative for support. Since we know that every site doesn’t has a live person to solve our issues, we find that most of the times you interact with a chatbot. The role of these things is to provide the customer w'ith information taken from different websites. Chatbots have advanced a lot. They have started understanding the problems of the users better and provide truly good results, which has been made possible because of a MLAs approach.

Search Engine Result Refining

There are numerous search portals like Google that uses ML to improve search outcomes. Every time a person performs a search, the ML keeps a track of our searching history and how one replies to the results of the search. For example, if someone has opened some of best results and is on the website for a long time then the search engine thinks that the result which it has displayed is right for the searched problem. Similarly, if you keep on going to subsequent pages w'ithout opening previous ones then then the search engine thinks that the results didn’t match the needs of the person and are inappropriate for the searched problem.

Online Fraud Detection

ML can also solve online fraud by always improving its software through constant evaluation of tasks. It shows its full power when making the Internet a completely protected platform. For example, ML is used by a lot of payment apps like PayPal for privacy and money transaction protection. The organizations use a lot of equipment to enable them to see and provide a comparison between numerous money transfers that are happening and distinguish among the various transactions which are recognized or unrecognized among different people.

Big Data Using Machine Learning

In today’s rapid, growing environment the amount of data gathered by companies is increasing daily. And even though the volume of data that is being collected is not important, the most important task is how the companies are using such a huge amount of data and making profit from it. The streaming of both structured and unstructured data from every corner of the globe at an enormous rate, establishing connections, and retrieving insights is a very complicated task that can swiftly spiral out of control. ML is based on algorithms that can learn from facts without depending on rules-based programming. Big data is the form of data that may be passed into the analytical system so that a ML model could “memorize” (or in other words, improve the accuracy of its prognostics).We often talk about ML and big data in the same flow' but they are not the same things. ML is required to extract the best information out of big data.

Enter Machine Learning

Modern businesses know that big data is powerful, but they’re starting to realize that it’s not nearly as useful as w'hen it’s paired with intelligent automation. With the increase in demand for ML, soon the price of AI will decrease, helping the whole world to adopt AI. AI machines have to be trained so that they can easily walk and communicate with people from different cultures and backgrounds. With massive computational power, ML systems help companies manage, analyze, and use their data far more successfully than ever before. Here’s how' multinational companies across industries are using big data technology to grow long-term business value.

ML and Big Data—Real-World Applications

Machine learning: The branch of AI that gave us self-driving cars has been useful in analyzing bigger, very complexed data to see invisible patterns, explore markets, and analyze the preferences of the customer for faster, more accurate results. In big data, ML is providing an interconnection of machines with huge databases to make them learn a lot of new things all by themselves. Big companies are analyzing big data with the help of MLAs to predict future trends in the market.

Healthcare: ML capabilities are impacting healthcare in profound ways, by improving diagnostics and personalizing treatment plans. Predictive analysis enables doctors and clinicians to focus on providing better service and patient care, creating a proactive framework for addressing patient needs before they are sick. Big data and ML help us to see the signs of disease, w'hich helps us to identify problems rapidly. It will also help in developing new' medications. The management of medical data, such as previous health records, reports from labs, etc., also becomes easier with the help of this. In return the data provides a clear view' of the health status of any person. Wearable technologies and sensors are also available to assess the patient’s health in real time, detecting trends or red flags that could potentially foresee a dangerous health event such as cardiac arrest. Advancements in cognitive automation can support a diagnosis by quickly analyzing large volumes of medical and healthcare data, identifying patterns, and connecting the dots to enhance treatment and care.fP. Y. Wu, C. W. Cheng, C. D. Kaddi, J. Venugopalan, R. Hoffman, M. D. Wang, 2017).

Retail: In retail, relationship-building is critical for success. The ML-powered technologies collect, examine, and work on that data to simplify the experiences of real-time shopping. The algorithms uncover similarities and differences in customer data to accelerate and make the segmentation simpler for better targeting. ML may have helped improve the accuracy of the workforce, but this is not entirely true. Since it is very true that the machines lack emotions and don’t feel any sentiments, they always require human involvement that can check the market conditions in different ways. Machines are only able to work as well as the algorithms are designed by us. However, based on learned preferences, deeper analysis can push undecided shoppers toward conversion. For example, ML abilities can help customers who are shopping online by providing proper recommendations of necessary products with reasonable prices, vouchers, and other real-time offers. With customer experience top of mind, Walmart is working to develop its own proprietary ML and AI technologies. In March of 2017, the retail chain opened Store №8 in Silicon Valley, a dedicated space and incubator for developing technologies that will enable stores to remain competitive in the next five to ten years.

Financial services: In the financial sector, predictive analytics help prevent fraud by analyzing large historical data sets and building forecasts based on previous data. ML models learn behavior patterns and then — with little human interaction — anticipate events for more informed decision-making. Now, by using big data and machine learning, we can reach the world trade market without distress. Banks and financial institutions use ML to gather real-time insights that help drive investment strategies and other time-sensitive business opportunities.

Automotive and other industries: In the face of stiff competition, the automotive industries are starting to leverage ML strength and big data analytics to make operations, marketing, and customer experience better before, during, and after the purchase of products. Applying statistical models to historical data helps automakers identify the impact of past marketing efforts to define future strategies for improved return on investment. This method of prediction helps producers and dealers check and explore sensitive data in regard to part failures, minimizing the cost of maintenance for customers. The network of dealers can be optimized by tracking for accurate, real-time parts inventory and improved customer support. As ML technologies hit new levels of maturity, smart businesses are shifting their approaches to big data. Across industries, companies are reshaping their infrastructures to maximize intelligent automation, integrating their data with smart technologies to improve not only productivity, but also their ability to better cater to their customers.

Implementing Machine Learning in Big Data

MLAs provide efficient, self-sufficient tools for data storage, analysis, and integration. If one’s organization is not very big and there’s not a lot of information coming it. ML wouldn’t be required because all the tasks could be done using a basic set of tools or manually. On the other hand, when one has an organization that deals with huge data, MLAs help to utilize time efficiently and effectively. Together with cloud computing benefits, ML allows rapid and precise analysis and summation of many datasets, whether they concern user behavior, sales, or DNA sequencing. Machines learn better the more data they have at their disposal. Big data analytics thus gives machines the volume and variety of data they need to make increasingly better and more efficient decisions in the performance of tasks. It makes sense — a veteran basketball player with a bigger and more varied “dataset” of experience will usually play better than a rookie.

The MLA could be implemented with many forms of big data operations:

  • • Data labeling/segmentation
  • • Data analytics
  • • Descriptive
  • • Diagnostic
  • • Predictive
  • • Prescriptive
  • • Planning
  • • Scenario simulation

Combining these elements allows users to see the big picture, created from big data with patterns, insights, and all other items of interest sorted out, categorized, and packaged into a digestible form. It’s important to understand that ML applied to big data results in an infinite loop. The creation of certain algorithms for specific tasks is being watched and upgraded over time as the data comes and goes through the machine (Volodymyr Bilyk, 2019).

Empowering Big Data and Machine Learning

As the world grows daily at an enormous rate, the size of data, collectively called big data, also grows exponentially. At the same time, another revolution taking the world to another level is going on in the field of technological enhancement, that is ML and AI. We can say in simple language that ML is a collection of equipment that empowers linked machinery and computers to learn, evolve, and improve by restating and frequently checking stored data and by continuously investigating human development. As far as the reality is concerned, technical giants, organizations, and data scientists all over the world are working on big data to make a big difference in the ML and AI world. In a recent survey conducted for big data executives by New Vantage Partners, about 88.5% of top executives believe that AI is very soon going to be seen as the biggest influence that might create a threat for their companies (Stevenn Hansen, 2019) (Figure 13.10).

Empowerment of big data w'ith ML

FIGURE 13.10 Empowerment of big data w'ith ML.

(Image Courtesy: Whatsthebigdata)

Conclusion and the Future of Big Data Analytics

As data sets continue to grow and produce more real-time streaming data, businesses are turning to alternative storage option like cloud storage, etc. The continuous exchange of data is changing the way we live in society. The huge impact of this bulk amount of data cannot be passed over. The amount of progress and developing changes is having both a direct and indirect impact on us; these new' discoveries are going to lead to huge amounts of data in the near future and managing this data will also become necessary. Since the powerful and accurate evaluation of data has provided us with the power and the ability to make decision in businesses, now more developed systems are needed to fulfill the requirements of future. The MLA provide huge support to handle big data. Numerous businesses are becoming knowledgeable about the difficulties of working w'ith a wide range of information. It is quite clear that big data and ML are going to be very important in the near future and it must be handled very carefully so that it can be easily used by anyone.


Volodymyr Bilyk. Oct 25. 2019. The App Solutions, development/machine-learning-and-big-data/.

Gunjan Dogra. May 15. 2018. Characteristics of Big Data. Indian National Interest, Sep28, 2019, algorithms- c33ef8488638.

Stevenn Hansen. Oct 14. 2019, Big Data Empowering, Oct 26. 2019, how-big-data-is-empow'ering-ai-and-machine-learning-4e93al004c8f.

Cynthia Harvey. June 5, 2017, Big Data Challenges, Datamation. Oct 25. 2019. https://www.

Shweta Iyer, Oct 1, 2019, Big Data Revolution, knowledgehut, Oct 26, 2019, https://www.

Matthew Mayo, May, 2018. ML Steps, KDnuggets, Oct 10, 2019 https://www.kdnuggets. com/2018/05/general-approaches-machine-learning-process.html.

Shweta Mittal. Om Prakash Sangwan, “Big Data Analytics using Machine Learning Techniques,” 9th International Conference on Cloud Computing, Data Science & Engineering (Confluence), Noida, India, 2019, pp. 203-207, doi: 10.1109/ CONFLUENCE. 2019.8776614.

J. Qui, Q. Wu. G. Ding, Y. Xu, S. Feng, “A survey of machine learning for big data processing”. EURASIP Journal on Advances in Signal Processing, vol. 2016, no. 67, pp. 1-16, 2016.

P. Y. Wu, C. W. Cheng. C. D. Kaddi. J. Venugopalan. R. Hoffman, M. D. Wang. “ML and big data-real world applications ‘omic and electronic health record big data analytics for precision medicine’”, IEEE Transactions on Biomedical Engineering, vol. 64. no. 2, pp. 263-273,2017.

  • [1] Smart speakers: e.g., Amazon’s Echo Dot and Google’s Nest Mini • Voice assistants: e.g., Samsung’s Bixby, Apple’s Siri • Smartphone applications: e.g., Google Assistant
< Prev   CONTENTS   Source   Next >