Looking to machine learning for solutions
By Yunnie Marzuki March 9, 2017
- Machine learning helps Grab Indonesia with efficiency, proximity and safety
- Bank DBS Indonesia uses machine learning to answer frequently asked questions
MACHINE learning is becoming an increasingly crucial element in today’s business world with most industries using it in one form or another.
Machine learning is a type of artificial intelligence (AI) that provides computers with the ability to learn without being explicitly programmed. Machine learning focuses on the development of computer programmes that can change when exposed to new data. The process of machine learning is similar to that of data mining.
Machine learning tasks are typically classified into three broad categories, depending on the nature of the learning "signal" or "feedback" available to a learning system. These are:
Supervised learning: The computer is presented with example inputs and their desired outputs, given by a "teacher", and the goal is to learn a general rule that maps inputs to outputs.
Unsupervised learning: No labels are given to the learning algorithm, leaving it on its own to find structure in its input. Unsupervised learning can be a goal in itself (discovering hidden patterns in data) or a means towards an end (feature learning).
Reinforcement learning: A computer programme interacts with a dynamic environment in which it must perform a certain goal (such as driving a vehicle or playing a game against an opponent). The programme is provided feedback in terms of rewards and punishments as it navigates its problem space.
The implementation of machine learning in industries such as banking, online services, and telecommunications vary tremendously.
At a panel discussion at CTI IT Infrastructure Summit 2017 in Jakarta on March 8, Grab Indonesia, PT. Bank DBS Indonesia and DataSpark discussed how they each utilised machine learning.
For Grab, the usage of machine learning is essential as it helps with efficiency, proximity and safety.
“In terms of the logic and algorithms, we put that into an account and learn the patterns of our drivers and customers from different cities. With machine learning, we are able to process our services in a better way,” said Grab Indonesia managing director of Ridzki Kramadibrata.
“As for drivers, machine learning helps them reduce costs as we known their patterns and it shows the productivity of drivers by learning how often they accept or reject orders,” he added.
Bank DBS Indonesia is transforming from a traditional bank into a digital bank with machine learning being utilised in their customer services department.
“We use machine learning as a virtual assistant. About 80% to 90% of our customers ask the same questions. We tutored the machine to deal with more ‘conversational language’ and answer these common questions,” said Bank DBS Indonesia head of digital banking Leonardo Koesmanto.
Currently Bank DBS Indonesia is exploring how machine learning can help to process documents based on customer patterns.
DataSpark is a new company under Singtel that provide business and government agencies with Singtel’s anonymised geolocation data.
“In DataSpark, we are using all the data from telco networks. One of our specialisations is to look at mobility signals in large areas so we apply both supervised and unsupervised machine learning algorithms to systematically look for certain patterns of our customers,” said DataSpark chief operating office Ying Shao Wei.
These three different industries face similar issues in implementing machine learning within their respective systems.
Grab Indonesia, Bank DBS Indonesia and DataSpark are all finding it difficult to find enough talent in the form of operators and supervisors to oversee the machine learning process.
Machine learning also has to work hand-in-hand with other technologies, functions, regulations and people, especially customers.
Machine learning also needs to become more conversational and the ability to obtain, process, and clean big data is not easy especially for big companies like Bank DBS Indonesia and DataSpark.
To overcome these challenges everyone in these industries has to work towards learning and understanding machine learning.
According to Ridzki, there is a possibility of connecting many industries in the business world. For example, there is a need to understand the patterns of customers from different industries
“We could pick up more important data that could help us in predicting businesses such as figuring out the number of people who are connected using a certain app,” said Ying.
“In banking, we can make connections for investors from customer patterns that are predicted based on data,” said Leonardo.