Machine learning: 6 innovative use cases in business today

Machine learning machine_learning_innovative_use_cases

 

Machine learning has rapidly shifted from an abstract idea to critical component for business intelligence.

Using sophisticated algorithms to identify patterns and discover greater insights from big data faster than any human, machine learning software learns without explicit programming and has already optimised and transformed several business operations all over the world.

There’s no better time to examine the many ways machine learning is applied in leading international enterprises and discover how it might be applied to your business operations. We breakdown 6 ways machine learning is helping businesses big and small innovate their products and services today.

 

1. Amazon - Product recommendations and fraud detection

Amazon is one of the biggest online retail stores in the world backed by the industry-leading Amazon Web Services cloud platform and high-profile artificial intelligence (AI) initiatives, so it’s no surprise the company already uses machine learning in key areas of its sprawling business.

The global e-commerce site has documented their ML solutions extensively, revealing it uses it in everything from pricing, product recommendations, demand forecasting and fraud detection to improve customer experiences. Amazon just made over US$51 billion in Q1 online sales, so we’re inclined to believe it’s making a real difference.

On their AWS cloud platform, machine learning models are deployed extensively in Glue and Macie, which use them to scan for patterns in data breaches for more efficient data cleansing.

That’s not even going into their consumer smart speaker products like Amazon Alexa & Echo, which utilise ML and AI algorithms to power its voice recognition and virtual assistant capabilities.

The reality is if you’ve had a chat with Amazon customer service or worked within AWS, you’ve most likely interacted with one of their machine learning models - whether you realise it or not.

 

2. National Australia Bank - Credit card applications

National Australia Bank recently announced they had discovered “hundreds” of use cases for machine learning in their business, implementing ML algorithms and natural language processing into their QuickBiz lending platform to create loan decision products for customers.

NAB’s ML initiative is powered by Discovery Cloud, an elastic data lake solution from Amazon Web Services. The bank currently leverages the platform to deploy its ML algorithms and natural language processing techniques into QuickBiz, which helps determine the success or failure of credit applications.

How does it work? By examining the text detail of transaction records for insightful patterns.

“We used a gradient-boosted model because it provided the best mix between the performance of the model to help us separate high risk customers from low risk customers, and the interpretability of the model.

"So in other words the model allowed us to understand why an application scored high or low, and also provided some additional predictive power which, at the end of the day, allowed us to approve more customers whilst still staying within the overall bank risk profile.”

- Kyle McNamara, at General Manager NAB

QuickBiz is reportedly one of the highest-rated digital services based on customer feedback, and while it’s still an “early” ML project for NAB, the value it gained from implementing ML is already clear.

 

3. Drive.ai - Autonomous Cars and Deep Learning

Machine learning AI in self-driving cars

 

The 2004 Hollywood film I, Robot depicted an intriguing future where our AI companions drove our cars for us, keeping our roads cleaner, faster and safer in the process.

While it may still seem like a far-flung concept, American company Drive.ai is bringing self-driving cars one step closer to reality by using the capabilities of machine learning and deep learning (DL) to enable future smart vehicles to be “adaptive and scalable” on the road.

DL is a sub-type of ML and structures algorithms in layers to make an artificial neural network that emulates the most human-like decision-making. Drive.ai’s team consists of several AI engineers out of Stanford University who built DL into every aspect of its self-driving cars.

The team transforms unstructured data from raw images into structured data, and use bounding boxes to elucidate common road objects like traffic light crossings and pedestrians.

The idea is the deep learning algorithms powering their self-learning cars intuitively recognise each element and figure out the best response while driving on the road.

 

4. Tiliter Technology - Smart Checkouts

Over 325,00 self-service checkouts will be deployed in Australian supermarkets by 2019 to reduce costs and increase efficiency - despite around $4.5 billion worth of retail theft attributed to shoppers abusing them.

Australian start-up Tiliter Technology is targeting future checkout cheats with its new smart checkout technology, which uses an automated product recognition system powered by ML algorithms to automatically identify and categorise grocery products for increased accurately.

“It’s based on machine learning and artificial intelligence which has been taught to recognise different types of fruit and other products.

“The big value for supermarkets is removing the significant loss seen from people entering the wrong information when using self-service checkouts. However, customers will also benefit by not needing to search through menus trying to identify the items they are purchasing.”

- Chris Sampson, co-founder of Tiliter Technology

It also removes the need for bar-codes or for customers to enter any other information on their purchase - it’s so adaptive it can tell minute differences between very specific products within the same family like fruit.

Shopping giants like Coles and Woolworths are reportedly in talks to trial them later this year, so you’ll most likely be interacting with machine learning every time you duck out for groceries.

 

5. Yelp - Image Curation

American review conglomerate Yelp boasts 92% of its visitors make a purchase after visiting their site - sometimes, frequently or almost always - and hosts tens of millions of photos uploaded by members globally, so it’s no surprise they stay on top of their user experience.

What many people would be surprised to learn is how they’ve kept ahead of the game: Yelp deployed machine learning algorithms as early as 2015 in a bid to future-proof its image classification service and create semantic data about individual photographs.

Today, Yelp’s machine learning algorithms identifies tens of thousands of related images and organises them under a diverse range of classes, based on a sequence of patterns.

The level of accuracy and efficiency afforded by ML helps customers find the exact photos they’re searching for and Yelp’s human employees better compile and label them.

 

6. Colombian Family Welfare Institute - Micro-targeting

Machine learning’s capabilities to transform our operations go far beyond cooler technology or more efficient transactions, as one compelling case in Columbia has proven.

The Colombian Family Welfare Institute (ICBF) is a welfare organisation that provides protection for over 8 million children and families across Columbia. In 2016, the group employed a local analytics firm using IBM SPSS Modeler to provide them with micro-targeting and predictive analytics to improve aid distribution.

The result? ICBF successfully delivered nearly 30,000 emergency food rations and 5 million dietary supplements to 38,730 malnourished children living in the most poverty-stricken areas of Columbia, who needed the help the most.

 

How machine learning can help your business now

How machine learning benefits your business

 

It’s fair to say with the amount of high-profile use cases available that machine learning is not all just predictive analytics and surveys. Along with the aforementioned examples, there are several more ways it and other predictive analytics solutions can enhance your business applications and systems today.

Ad targeting: Advertising tailored around the needs of our customer’s behaviours and traits, with ML adapting what ads are shown based on their evolving needs and requirements.

Churn analysis: ML algorithms identify when and why customers are likely to leave before they do so, allowing you to target them with incentives and promotions to retain their business.

Equipment monitoring: Real-time information gathered by Internet of Things (IoT) devices equipped with sensors can be examined by ML software to predict when a device might need maintenance or updates.

Spam filtering: Machine learning algorithms can identify patterns in spam messaging and automatically filter them out without explicit instruction from the user.

 

Related posts