4 Unexpected Uses of Machine Learning

4 Unexpected Uses of Machine Learning

The pace of technology is continually accelerating and we are seeing more and more machine learning solutions being utilized across industries and even in our personal lives. But beyond the practical application of the technology, data scientists are developing some truly amazing use cases for how ML can enhance our lives unexpected ways. From identifying which shop your ramen noodles came from to saving the rainforests, the possibilities are endless with this innovative technology!

Creating New Music Sounds with ML

NSynth (Neural Synthesizer) is a machine learning algorithm that creates new music sounds. Google’s Magenta uses this algorithm by understanding the components of musical sound to develop completely new sounds through deep neural networks. Magenta’s prototype was developed with 16 sounds and 15 different pitches- this resulted in 100,000 new sounds. Google has created NSynth Super for artists to access and utilize the open source ML to have the ability to create new music. With the unpredictability of machine learning outputs, it’s possible this ML algorithm can even make a completely new genre of music!

ML Can Predict How Much Medication You Need

Dana Lewis’s OpenAPS (open artificial pancreas system) is a ML solution for individuals with diabetes. The OpenAPS provides a predictive algorithm that gives diabetic patients personalized recommendations and actions for treatment. All relevant data points are connected to automatically allow the correct amount of insulin to be released. The OpenAPS has proven to be a more effective treatment for diabetics than the current insulin pump therapy.

Where Are Your Ramen Noodles From?

You can now use ML to figure out what shop your ramen noodles came from in Japan! With the use of AutoML Vision and machine learning models, data scientist Kenji Doi created a solution to figure out what store each ramen noodle dish came from. Kenji uploaded 48,000 photos of ramen dishes to AutoML Vision and the model was trained within 24 hours. To test the ML, Kenji submitted photos of various ramen dishes to AutoML Vision and the solution was 94.5% accurate in predicting which shop the noodles came from. Learn more here.

Saving Rainforests with TensorFlow

Google has built an alert system to conserve rainforests. This ML solution uses audio and TensorFlow to detect sounds of chainsaws and logging trucks to understand if any if an illegal activity is occurring. The ML takes into account various AI techniques to ensure it is correctly detecting any destruction taking place. Combating deforestation has always been a challenge and technology is a solution that actually works in fighting this issue.

To learn more about how what’s possible with machine learning, contact us here.

By | 2018-04-30T15:54:15+00:00 April 30, 2018|Categories: Fusion Blog, Innovation Posts|

About the Author:

Klaudia Gatowska
Klaudia Gatowska serves as the Marketing Associate Manager at Maven Wave. Ms. Gatowska specializes in strategic marketing encompassing advertising, social media and digital platforms. Prior to Maven Wave, Ms. Gatowska served as a Marketing Intern at The I Hotel and Conference Center following a Marketing Internship at Maven Wave during her senior year of college at The University of Illinois Urbana-Champaign.