Machine learning has crossed from the lab to the business world. Machine learning provides insights that help to create more intelligent data-driven applications that improve business processes, operation, and easier decision making.
In a conversation at Structure Data 2016 conference in San Francisco, Dr. Peter Lee, Corporate Vice President, Microsoft Research and Jack Clark, Bloomberg News – San Francisco, talked about the advances we made in Artificial Intelligence (AI) and machine learning in recent years. Dr. Lee is responsible for Microsoft Research New Experiences and Technologies. He said that AI is essentially used to really understand what customers want. For example, machine learning software tools ‘get better’ the more people use them, since the algorithms ‘learn’ the user’s behavior. To get insights that enable meaningful and measurable improvements, artificial intelligence (AI) experiment work needs huge quantities of data, coupled with super fast computing power and cloud technologies. Dr. Lee said that machine learning takes us further; it brings the premise of reinventing new ways to solve challenges.
Strategists and developers point out that AI seems scary to many people outside this field and a common connotation is that “the machines are taking over us, humans”. But, when experimenting with an AI model, the unpredictable results are actually fun. One example is speech-based apps that learn to adjust to the individual accent, pronunciation, and dictation nuances (i.e. Does Siri always gets it right?)
In speech processing in real life, people interpret body language, facial expressions as well as tone of voice. AI can learn to differentiate between a sincere compliment or sarcasm, for example, when a person says ‘I really appreciate you’, or ‘This coffee tastes good’. The ability to create an intelligent speech model that has deep learning capabilities is doable with today’s computing, algorithm development, and sensor technologies.
Machine learning can provide us with valuable information in public and personal health. In healthcare, the fundamental challenge is to aggregate all the medical records into one huge data repository. Today, even a stand alone health facility doesn’t have visibility to how many patients are being treated for a specific condition and which treatment options were offered or were selected. Before any machine learning can be applied, getting the data from all the health-related pipelines is a basic tenet. However, once we overcome this challenge, AI models can be used to help in many ways, such as in diagnosis of conditions, looking at the genetics make up of the patient and offering treatment options, assessing geographical spread of infectious diseases, and more.
Intersecting user data, operational and tactical data, knowledge management, analytics, and machine learning, coupled with the ability to analyze these combined resources and provide context, are the new growth phases in data science.