How Machine Learning & Artificial Intelligence(ML/AI) benefits top MNC's? Amazon Use-Case

We all know about the tech giant Amazon Inc. It is an American multinational technology company, founded by Jeff Bezos, which focuses on e-commerce, cloud-computing, digital streaming, and artificial intelligence. It is considered one of the Big Five companies in the information technology industry, along with Google, Apple, Microsoft, and Facebook. Amazon is also referred as "most valuable brand". Amazon also provides Cloud Computing service named "Amazon Web Service". It implemented the several technologies to work together so that the user get seamless experience. Behind the scene of providing the great user experience, one of the key concept used by Amazon is Machine Learning.

What is Machine Learning (ML)?

When I googled about Machine Learning, I literally got scared with the tough vocabularies used there. But here I'll try to explain Machine Learning in the most simple language. So basically, "Machine Learning is a part of Artificial Intelligence, it is the concept using which we train our computer systems i.e. we try to provide intelligence to the computer system, so that it also become capable of evaluating and predicting futuristic data".

To explain this I am giving you a small example, so that it becomes easy to relate and understand about historical data, futuristic data, etc. Example:

I guess you were able to find the value easily which is '80', right? Now, the crazy thing is that the concept or the algorithm(steps) which you used just now to find out the answer is known as the term called 'Intelligence', you used nothing other than a simple mathematics, your brain and the data which was already available in the table to find out the answer. The answer which you gave, i.e. '80', without providing to you is known as Futuristic data and the table you used(in your sub-conscious mind) to find the answer is known as Historical Data. And when we embed this intelligence into computer system or machine using some programs and historical data, we are actually making our machine to learn these concepts so that it also become capable of predicting future data, this process of making machine learn these things is known as Machine Learning.

Machine Learning uses a very great concept known as LinearRegression().

It uses a very basic formula i.e. : y = b + c*x, where y is target variable, b is bias data, c is coefficient and x is predictor.

You might be thinking that it was so easy then why we need computer to learn these concepts? So the answer to this is pretty simple, in the real world, we use machine learning to predict much complex things, the things which human mind is somehow capable of finding the solution but it will take a lot of time and here the machine will win, like we use Machine learning to predict the weather forecast, predict the chances of winning team in cricket match, predict arrival of monsoon winds, predict the number of customer will buy a particular product, etc.

Why there is the need of Machine Learning?

Gone are the days when programmers would tell a machine how to solve a problem at hand. We are in the era where machines are left to solve problems, on their own, by identifying the patterns in each data set. Analyzing hidden trends and patterns makes it easy to predict future problems and prevent them from occurring. A Machine Learning algorithm usually follows a certain type of data and then uses the patterns hidden in that data to answer more questions.

"Humans and Machine Learning algorithm can only predict future data, but never ensure the information to be 100% correct."

Let's discuss some of the vocabularies used with Machine Learning and Artificial Intelligence (ML/AI).

Data Analysis vs Data Analytics

Most of the peoples are confused between these two somewhat similar sounding terms. So the most important difference between these terms is, When we do certain operation in our historical data or Dataset, like sum, diff, min, max, etc. then it is termed as Data Analysis while when we do certain operation so that we can predict the future data is known as Data Analytics. Both the terms are the part of Data Science. Like this in ML also we work on the Data set, so we can conclude that somewhere Machine Learning is also a subset of Data Science.

Artificial Intelligence

Artificial Intelligence (AI) is the branch of computer sciences and technologies that emphasizes the development of intelligence machines, thinking and working like humans. For example, speech recognition, problem-solving, learning and planning, etc. With the advancement in technology, we are already connected to AI in one way or the other whether it is Google Assistant, Siri, Cortana or Alexa. AI is gaining popularity at a quicker pace; influencing the way we live, the way we interact, and it also helps to improve user experience.

Are Machine Learning and AI same?

The answer is: No, Artificial Intelligence and Machine Learning are not same, but they are closely related. Machine learning is the method to train a computer to learn from its inputs (the data provided to it) but without explicit programming for every circumstance. We can say that Machine learning helps a computer to achieve Artificial Intelligence.

How Amazon uses Machine Learning and Artificial Intelligence?

How ever amazon looks as simple as an e-commerce cum entertainment platform, but behind the scene Amazon is very much reliable on using Machine Learning. From Advertisement, Supply Chain Optimization to Modeling and Optimization, Amazon is dependent on Machine Learning. One of the main areas where Amazon is applying continuous AI is to better understand their customer search queries and what is the reason they are looking for a particular product. For an e-commerce company to make relevant recommendations to its customers, it is not only crucial for them to know what their customers searched for, but it is also critical to understand why a customer is searching for a product. Understanding the context can help the retailer to recommend complementary items to its customers, and Amazon is intent to work out this puzzle by applying AI to the problem. Amazon also uses ML - to decide on what, when and how to send personalized offers/discount notifications; Promotion Effectiveness; Marketing Mix Analysis; Inventory/Demand forecasting, etc.

There are certain teams in Amazon service that substantially depend on Machine Learning, below is the words by some of the service team of Amazon who are using Machine Learning:-

• Advertising Technology - Amazon is investing heavily in building a world class advertising business, by solving challenging machine learning problems. We analyze billions of ad impressions and millions of clicks daily to solve various challenging problems related to robot detection, publisher quality, contextual extraction, relevance modeling, recommendation systems, performance prediction and impression/click pricing, and much more.
• Alexa Engine - Help us make Alexa personalized to each of our customers. Our mission is to apply Artificial Intelligence and Machine Learning, in order to reduce users’ cognitive load, reduce friction in their day-to-day activities and finally, inspire our customers by enabling serendipitous discovery of experience.
• Alexa Science - Alexa is the cloud-based intelligent agent that powers Echo and other devices designed around voice. We are a team of scientists and engineers creating the technology behind Alexa. Our mission is to push the envelope in AI, NLU, ML, Dialogue Management, ASR, TTS and Audio Signal Processing in order to provide the best experience for our customers.
• Alexa Smart home Devices - Alexa Smart Home is focused on keeping customers safe by detecting unexpected behavior, saving money by managing energy use, and saving time by learning household routines. Our teams are researching the next-generation approaches for Alexa to continue learning about users' routines and execute actions based on those routines, resulting from an event trigger (e.g. a customer waking up, or coming home from work). Our scientists will be focusing developing the internal machine learning needed to create that learning process for customer routines, and continuing to test new methods to improve that system
• Amazon JIHM - This is the team that has just launched the Amazon Go stores and service using state of the art Computer Vision technology. This team continues to expand on the initial model they created with a focus on expanded machine learning utilization and improving efficiencies. We’re also considering new ideas and concepts in Computer Vision that can help take Amazon Go to the next level.
• Amazon Music - Our San Francisco based team owns Amazon’s core digital music services. Our music offerings are available in multiple countries, and our research applications support the mission of delivering music to customers in a way that enhances their day-to-day lives.

• Amazon Rekognition- Amazon Rekognition is a Deep Learning-based image recognition service that enables developers to search, verify and organize millions of images. We are a team of scientists and engineers creating the technology behind the Amazon Rekognition cloud service. You will work on state-of-the-art Deep Learning and NLP techniques to create scalable solutions for non-trivial, arguably unsolved problems in computer vision.
• Amazon Visual Search - Visual Search technology helps customers use visual information for search, discovery and shopping. We create Augmented Reality solutions on mobile devices, overlaying relevant information over camera-phone views of the world around us powers solutions that lets customers search for products based on their visual attributes such as color, shape or even texture. Such solutions appear on Amazon and Zappos, allowing customers to quickly find the shoes or watches they like based on the appearance of the product.
• Customer Service Personalization - We use Machine Learning, NLP and Statistics to provide the best customer experience on the earth. Our team is building the next generation of intelligent customer service. Join us to build revolutionary products and change the way that people work with customer service.
• Modeling and Optimization(MOP) - We build deterministic and stochastic models and algorithms to grow and optimize Amazon’s transportation network. Challenging problems include ideal location and capacities for installing Prime Air locations, Sortation centers, delivery stations, Amazon Lockers, Prime Now and Fresh Nodes for multiple years into the future.

• Supply Chain Optimization- SCOT builds algorithms that manage the inventory movement within the Amazon supply chain. This includes demand forecasting for tens of millions of SKUs per day, multi-echelon inventory optimization, strategic procurement, capacity planning, simulation, experimentation, and inventory placement. They optimize tradeoffs among customer promise, fast track delivery, sourcing cost, transportation cost, holding cost, inventory obsolescence, fulfillment center capacity and location. They automate and optimize Amazon’s supply chain to support the three pillars of Amazon’s Consumer business: price, selection, and convenience.
• Voice based and Advanced Shopping- We strive to enable shopping in everyday life. We allow customers to instantly order whatever they need, by simply interacting with their smart devices such as Echo, Fire TV, and beyond.

______________________________THANK YOU______________________________

Aditya Kumar Sahu, Birla Institute Of Technology, Mesra, Ranchi

#ARTH task 5.