The Importance of Machine Learning

by Marco Tapia

in management-it-consulting,

April 7, 2017


The Importance of Machine Learning

It's been a few years that machine learning has been used to a large extent for image, text recognition and video. Today however the uses have increased and we see it being used in cyber security, improve medical outcomes and business applications to name a few. Machine learning today isn't like any machine learning of the past, because the ability to automatically apply complex mathematical calculations to very large data sources over and over again at tremendous speeds- is a recent development and one that is gaining momentum in the science/computer worlds. (1)
Machine learning today makes it possible to quickly and automatically produce models that can analyse bigger, more complex data. This will also deliver faster, and more accurate results. By building precise models, an organisation has a better chance of identifying profitable opportunities and safer outcomes – or avoiding unknown risks.(2)
Machine learning can be applied in cases where the desired outcome is known called (guided learning), or the data is unknown (unguided learning), or it can be used if learning is the result of interaction between the environment and a model called (reinforcement leaning). (2) Importantly it removes human intervention and becomes machine orientated by thinking all on its own.
So let's look at who is using machine learning in everyday applications.

"Who is using Machine Learning"
The chief executive of accounting software business Xero claims machine learning-based automation will be a huge asset and will change the advent of cloud computing as it commences to 'automate' accounting tasks of business. (3)
For small business this will remove essentially what has been business owners/workers or accountants inputs into daily business transactions to now seeing the automation of coding invoices and bank transactions. Xero boss Rod Drury, believes that this application is the first of its kind in the industry which will now spawn off to a new wave of innovation for accounting. (3)
To read more about Xero click here.
Let's look at the the new IBM Watson project. IBM Watson has always been about putting the power of data science into the hands of the masses. Last year, IBM announced another step towards that vision with the launch of the "Watson Data Platform".
"The theory is simple. The incredible potential for driving efficiency and change with Big data and advanced analytics - as well as all the associated technologies such as machine learning, the Internet of Things, and predictive modelling - is so great, it should be available to everyone. Not just those who have spent years in college studying the fundamental mathematical and statistical systems under the hood of today’s analytics toolsets" (4).
In fact, IBM say that the Watson platform is the first enterprise data platform built from the ground up to enable machine learning – as Rob Thomas, vice president of product and development for analytics, puts it, “steeped in artificial intelligence.”
“For the first time,” Thomas is quoted as saying, “You can bring all your data to one place and it’s immediately catalogued and organised and ready to apply artificial intelligence and machine learning". To read further about IBM Watson click here.
Lisa Morgan from Information Week has put together an interesting read on the "11 Coolest Ways to use Machine Learning". Here are some applications which are already in the process and developing into more powerful automated intelligence. (5)
"11 Coolest Ways to Use Machine Learning"
1) Malware
In 2014, Kapersky reported it was detecting 325,000 new malicious files every day. At that rate, humans and even signature-based security solutions could not keep up, which is why machine learning and deep learning are necessary.
Deep Instinct uses a large core of several million malicious files, tens of millions of legitimate files and malware that Deep Instinct may have mutated by 20% - 50% for training purposes. The more radical malware mutations make the training more difficult, but they also make the model more resilient. Once the training is finished and the synapses have been updated, a text file of the synapses can run deep learning in prediction mode.
2) Make Important Discoveries
The healthcare industry is constantly looking for ways to prevent diabetes and minimise its effects.
Medecision used a machine learning platform to gain a better understanding of diabetic patients who are at risk for avoidable hospitalisation or emergency room use. The model identified seven or eight independent variables that can be used to predict avoidable hospitalisations on 8 million patients. The surprising indicator was whether the patient had a flu vaccine. The analysis indicated that most of the avoidable hospital admissions were for upper respiratory infections that were complicated by diabetes but not caused by it!
3) Understand Legalese
Legal documents are often too complicated for the average person to easily comprehend. Some hire a lawyer. Others may skim the documents, or even ignore a document's content.
Legal Robot can determine what's missing from a contract and whether there are elements in a contract that shouldn't be there.
4) Prevent Money Laundering
PayPal is using deep learning to prevent fraud and money laundering. . By combining deep learning with machine learning and other tools, the company can precisely discern between legitimate and fraudulent buyers and sellers.
5) Improve Cyber Security
An Israeli communication service provider has been using machine learning for the past two years to help protect its business and customer data. The new system monitors all traffic coming from and being exchanged among PCs and servers, to identify anomalous behavior. Recently, the system detected malicious code in a video file that an employee had downloaded. The security team instantly notified the employee.
6) Compete Intelligently
Improving your position in the Tour de France is difficult if you have little or no perspective into the positions and status of other cyclists. About 200 cyclists participate in the race, and not all riders are covered on TV.
WinningAlgorithms is able to determine what is occurring in the race 5 mins before broadcast, therefore it helps place those riders on the TV stage more accurately. This has been used since 2012.
7) Get Ready for Smart Cars
The IBM Institute for Business Value surveyed 175 auto industry executives in 21 countries. Seventy-four percent expect that by 2025, vehicles will self-optimise and provide advice in context. Specifically, they'll be able to learn about themselves, the surrounding environment, and the behaviours of the drivers and the occupants.
Future vehicles will also be able to personalise driving experiences by observing and mimicking their human drivers/owners.
8) Mitigate eCommerce Fraud
Retailers employ analysts to help identify, reduce and prevent fraudulent transactions. Many have used rules to block transactions from suspicious locations, such as Nigeria and Ukraine, but that approach also blocks legitimate transactions. Machine learning helps retailers and others manage fraud in a more precise way.
The goal is to identify fraud patterns before a product ships, without delaying the delivery of products.
9) Fine-Tune Security Screening
Airline passengers, concert attendees, and sports fans have something in common: they're screened by security guards and systems. Humans often overlook items that machine learning can identify. And, machine learning can easily adapt to seasonal changes affecting bag types and bag contents, or the specific requirements of a particular venue.
10) Improve Customer Service
Machine learning can improve the efficiency of customer service by understanding customers and their issues at a granular level.
Machine learning can easily discern between the customers that are beginning to use a product versus those that have more experience with the product, which enables efficient customer support. Alternatively, it can recognise and proactively address customer issues as they occur.
11) Outsmart the litigator
Historically, lawyers and their staff have manually reviewed court documents, which can take weeks or months. Machine learning can speed the process and uncover important details humans may overlook.
Machine learning can look for patterns in language that indicate peoples' behaviour. It is this pattern of behaviour that is very important in court scenarios where the lawyer will be able to prove it more efficiently. (5)
Machine Learning is obviously a growing field and there should be many exciting times ahead. If you want to get involved have a look at this fun video from 'Hello World-Machine Learning Recipes' on how to code. Click here to watch video.