By Chai Hung Yin
An apple a day might train a machine one day.
We all know 'A' is for apple, but how does a child recognise that an apple is called an apple?
The answer to this question helps us understand the concept of artificial intelligence (AI).
“Showing him a picture of an apple just once is not enough,” said Ms Sze Chi Lim, a machine-learning engineer with Hanalytics, an AI company that specialises in using advanced machine learning technologies. “You have to do it repeatedly, show different kinds of information and characteristics of an apple and let the child learn,” she added.
Using a similar concept, Hanalytics assists doctors in diagnosis by detecting tumours as well blood vessel diseases, uncovers insights and creates predictive personalised applications in the healthcare sector. The year-old homegrown startup partners some of the best hospitals around Asia, including National Heart Centre Singapore, and through these hospitals, obtain top quality data and a high volume of patients.
“Our predictive models are built based on learning from the data collected from the hospitals. As the hospitals we collaborate with are the cream of the crop, large data quantity and high data quality can be obtained. This is how we predict not the best outcome, but the most accurate outcome,” said Hanalytics’ principal software engineer Junon Lim.
According to him, the company is able to do medical prediction based on CT and MRI scans.
Screenshot of Brendan Lim from DataRobot, from an IMDA video on AI.
But what exactly is AI?
One way to define AI: Think of it as referring to technologies that aim to simulate human processes or traits, including problem solving.
"Websites that recommend movies, (which are) tailored to your personal preferences, are another good example of AI,” said Mr Brendan Lim, regional sales vice president of US-based DataRobot, a company that offers machine-learning platforms for data scientists.
For a machine – or the movie website, in our example – to achieve intelligence, it needs to learn using data points. “Maybe it will rely on your age, country of residence, your previous movie rental purchases, or your ratings of various movies,” explained Brendan. “It is really about how we leverage data to make decisions. And machine learning is how we take that data and let the computer learn from that data.”
Brendan said small and medium-sized enterprises (SMEs) in the retail industry can use AI to predict inventory and forecast sales – one of the things DataRobot can do to help SMEs kickstart their machine learning journey by leveraging on their data and specific use cases.
AI's role in a Smart Nation
Mr Tan Kiat How, CE of IMDA.
AI has been identified as one of the four frontier technologies that are essential in realising the country’s Smart Nation vision through a Digital Economy. In fact, in May 2017, AI Singapore – a national initiative to develop deep AI capabilities – was set up and the government is pumping in up to S$150 million to “catalyse, synergise and boost Singapore’s AI capabilities”.
Growing these frontier technologies – including cybersecurity, immersive media and the Internet of Things – is one of the thrusts of the recently launched Infocomm Media Industry Transformation Map (ICM ITM), Singapore’s roadmap to powering its future Digital Economy.
To support the growth of companies in the Digital Economy – which includes AI-focused enterprises – IMDA also announced the rebranding of Accreditation@IMDA programme to Accreditation@SG Digital. DataRobot is one of the newly accredited AI companies under this programme.
Mr Tan Kiat How, Chief Executive, IMDA, said that it is important to spotlight AI because its immense potential. “(We are) using technology and machines to really improve efficiency in the workplace, create new business models, and create jobs that we cannot even imagine today,” he said.
Screenshot of an IMDA video on AI, illustrating how the Government can work with companies to use AI to make lives better for everyone.
As technology advances, he added, it is a continuous journey to remain relevant at the frontier, and most importantly, use technology to create competitive advantage for companies and good jobs for the people.
To catalyse AI adoption and create a pervasive AI ecosystem in the country, IMDA and AI Singapore have also launched an AI Business Partnership Programme to facilitate partnerships between local companies and AI solutions providers.
IMDA is also helping to train AI talent pool for industries via the AI Singapore Apprenticeship Programme; building AI libraries such as the National Speech Corpus; and helping SMEs assess if they are AI-ready through the SME Digital Tech Hub.
SMEs can visit their nearest SME Centres for more information on IMDA pre-approved technology solutions. For those with more advanced digital needs, including in areas such as data analytics and cybersecurity, the Business Advisors at the SME Centres will refer them to the SME Digital Tech Hub.
For more information, check out the SMEs Go Digital programme page.