How machine learning is changing the way we live and work
The mention of the term “Artificial Intelligence” in the media is almost always accompanied by dire warnings of a self-aware system that can wipe out humanity, enslave us or put us out of jobs.
No news article on this topic is complete without a reference to the Terminator movies, accompanied by a photo of a robot with an evil grin and red camera eyes looking at us. But maybe it is just happy to see us, so why be prejudiced against it!
Enterprise AI is all about cognitive computing
Date: Thursday, Sep 27, 2018 | Time: 10:30AM TO 11:15AM CT
Duration: 45 Minutes
- Basics of cognitive computing & AI
- Integrating cognitive computing APIs with enterprise systems
- Significance of cognitive computing & AI across industry verticals
- Use cases of cognitive computing combined with AI
- Present day examples/ case scenarios
- Q & A
This blog post is how machine learning, a subset of artificial intelligence, is changing entire industries, the way we work, our homes and improving our health.
Machine learning involves a software program which uses algorithms to process data. There are different kinds of algorithms for this purpose, but they all work towards the same goal – analyzing the datasets given to the program in order to learn from the data and make predictions for future behavior.
The two most common methods in machine learning are supervised learning and unsupervised learning. Supervised learning involves training the algorithm on a large set of labeled data and then used on other datasets to get predictions. Unsupervised learning involves letting the algorithm loose on unlabeled datasets to learn on its own so that it can recognize the patterns in the data. The other method is reinforcement learning, which involves a trial-and-error approach.
As the size and variety of data increases, the program’s algorithm keeps getting better and results in more accurate predictions. One of the most promising machine learning methods right now is deep learning, which involves the creation of artificial neural networks.
Here are some well-known examples of machine learning which are already a part of our lives or on the way to be one –
- Google’s self-driving cars
- Detecting credit card fraud
- Facial recognition in Google Photos and Facebook
- Apple Siri, Google Now, Amazon Echo and Windows Cortana
- Email providers using machine learning to detect and handle spam emails
- The recommendation engines used by Amazon and Netflix to show you items and movies based on your previous site interaction
Machine learning is making important contributions in these four fields currently:
Machine learning is making healthcare more accurate and saving lives
Hospitals can predict the emergency room wait times for patients by analyzing factors such as staffing levels, patient data, and ER department charts. This helps hospitals plan better and allocate resources based on predicted inflow.
IBM has developed a machine learning algorithm that can “read” the hand-written notes of physicians stored in electronic health records and carry out heart failure diagnosis just as a doctor would do by reading his colleague’s notes.
Healint, a Singapore-based company, has an app that helps patients suffering from chronic and neurological diseases such as strokes, migraines, and epilepsy. It works by using the smartphone’s sensors and algorithms to identify warning signs for neurological conditions. It also has an emergency alert system that patients can trigger just by shaking the smartphone and alerting their caregivers. The algorithms-driven app can differentiate between regular shaking and emergency shaking.
Microsoft’s ongoing Project Oxford is using the company’s Azure Machine Learning service to analyze facial images and identifying different human emotions. It is now able to identify if the individual in the photo is happy or sad. This has huge implications in the field of psychology and in providing care to people battling mood disorders.
Machine learning is also helping doctors monitor skin lesions, creating social AIs to help autistic kids feel less isolated and analyzing the data gathered from sensors put on surgical tools to replicate the perfect operation.
It is also helping medical researchers identify human genes that make people more likely to develop certain types of cancers and develop drugs that have a higher chance of success.
Even predictive maintenance of hospital equipment done with the help of machine learning software can help healthcare organizations deliver better patient care.
How AI and Data Science solve the biggest business challenges of today
- An Overview on Data Science and Artificial Intelligence
- Why every business needs Data Science and AI
- Business cases for varied industries
- Roadmap for implementation
Big retailers are using it to predict consumer behavior
Amazon is the most well-known retail industry user of machine learning. The company uses machine learning to provide the lowest prices possible compared to its competitors. It achieves this by using algorithms that make comparisons with the prices of same and similar products on sale by other online retailers.
Ralf Herbrich, Amazon’s director of machine learning, explained how this works out at a recent machine learning and big data conference in Berlin. Amazon sells about 20 million products and its success is in large part due to the fact that it is able to offer a lower price for almost any product compared to its competitors. This means that manually checking out competitor websites all the time and making changes to the prices in Amazon constantly is just not possible.
This is where machine learning and analytics comes in play. A predictive model goes out and tries to find the same or similar items for sale, including items that have a similar product description to determine prices in Amazon.
Just in time for this holiday shopping season, IBM has launched the Watson Trend App which is also available as an iOS app. Shoppers can use this app to know which products are trending this holiday season and also the reasons behind it. Currently, it is tracking products in three categories – toys, tech, and health.
This will give consumers an idea of what’s popular and buy those items before they are sold out. As the name suggests, IBM is banking on the popularity of its AI-driven Jeopardy champion to let consumers know how reliable the predictions are.
The Watson Trend app works by carrying out a sentiment analysis of millions of online conversations from over 10,000 sources which includes social media sites, forums, blog posts, comments, reviews, and ratings. It also shows what consumers are feeling about those items.
All this is based on a combination of Watson’s natural language capabilities, predictive analytics and a combination of API capabilities from the Watson Developer platform. The app gives users a daily trend score ranging from 1 to 100 based on the size of the conversation and the rate of growth of the conversation.
While this is being given for free to consumers, the business potential of this is huge. Any retailer would love to have the knowledge of what consumers are planning to buy around the year and what they are saying about a particular product or service.
Helping farmers get weed-free farms
Robots are going to play a huge role in the automation of farm work and machine learning is going to be an important part in this field (pun intended) as well.
BoniRob is a robot made by Bosch’s Deepfield Robotics division along with the agriculture department at Osnabrück University and AMAZONE H. Dreyer, a German manufacturer of agricultural equipment.
Among a variety of farm tasks that this robot can carry out, one of them is weed removal. A bolt-on module uses machine learning to recognize weeds and kill them using a small punching tool that pounds them into the ground and even perform herbicide at the rate of 2 weeds per second. This ensures that no unnecessary chemicals are used on the farm so that the nutrients are left in the soil. This technology helps farmers maintain soil quality.
Enterprises are using machine learning to forecast demand for their products
It’s not just retailers who are using machine learning for predicting demand. Even B2B companies are using machine learning to know which products are in demand in real-time so that their dealers are delivered the right amount and type of products.
Cisco Systems is one best-known users of machine learning in this regard. It has a collection of 60,000 propensity to buy (P2B) models which it uses to forecast demand for its entire product range. It includes every potential customer in every location in the world. Cisco’s sales and marketing teams rely heavily on this model in their day to day work. The company uses a variety of tools, including machine learning, to keep these P2B models up-to-date every quarter. This includes retraining them and ensuring that the sales and marketing staff get access to it before the data grows stale.
There are several other uses of machine learning that are in use across many industries right now. Some of these are:
- Detecting bank fraud
- Factory maintenance diagnostics
- Categorizing images, video and text
- Automatic speech recognition used in converting speech to text
- Automatic packaging plants, e.g., sorting and packing different kinds of fish into different cans automatically
Companies have started realizing the benefits of machine learning in automating routine, mundane tasks and tasks which require a huge team of people, tasks involving Big Data and requiring quick, real-time action. To learn more about how machine learning can help your business, talk to our data scientists for a detailed use case.