This article was written by Joshua Chan, an intern here at Hackwagon Academy.

 

This summer, I took the plunge and decided to intern at a tech start-up. A psychology major, I was the proverbial fish out of water at Hackwagon Academy, a programming school for adults. Save for some basic HTML I learnt for blogging when I was 12/13 (as one does) and some research done about Big Data for school projects, I had but a vague concept of Data Science. I had seen words like ‘Internet of Things’ and ‘Machine Learning’ thrown around a lot but I never really understood their meanings. I was also aware that the government was paying much attention to these ‘things’ but I chose to remain oblivious to them. That, however, clearly did not bode well when I was tasked to run a marketing campaign for Hackwagon’s Data Science bootcamps. Alas, so began a crash course of my own as I endeavoured to understand what the data science hype was about.

My colleagues were, fortunately, patient in unpacking tech terms for me and explaining current trends. As such, in the short 4 weeks since the start of my stint, I can say that I have gotten a reasonable grasp of this subject. Granted while I’m nowhere near an authority on the subject, I thought it would be nevertheless interesting if I could share my understanding of how and why ‘data science’, mysterious the two words they are, have become the buzzwords of today.

Trusting that my tech lexicon is not all that well developed yet, I would imagine this becomes a simple and digestible primer, if you will, on the importance of data science. With that, I hope readers who are not acquainted well with ICT and programming could better understand the phenomenon of data science in Singapore.

 1: Artificial Intelligence and Machine Learning

AI is one branch of data science that has long caught the attention of many fascinated with technology and in 2017,the Singapore government pledged to invest a majority of $150million into AI to boost its financial, city management and healthcare solutions. Given the apparent importance of AI, it would perhaps be good to understand how data is involved in the creation of self-sustaining intelligent systems.

How is data processed in a way that allows the intelligent non-humans we see in sci-fi films to come to be?

To answer that, one looks at machine learning (ML), a commonly acknowledged adjunct to AI.

Without going into technicalities, ML simply put is feeding data into an algorithm, which adapts to its own mistakes to give accurate predictions and outcomes. A relevant example for this is the automatic Facebook Tagging function. When one first tags a friend, information about the coloured pixels that make up his or her face are then fed into Facebook’s algorithm to predict future tags. Sometimes, Facebook mistakenly tags people or misses out on a tag altogether. In that case, if the user corrects for these, the algorithm is adjusted accordingly to make better predictions in the future.

The system by which this algorithm evolves is Deep Learning. As to how this works, that’s for another time. Insofar as ML is concerned, one could probably draw links to how systems execute their functions precisely and accurately.

2: Cybersecurity

Another ICT domain with which news have frequently talked about is Singapore’s push to develop its cyber-security capabilities. This year alone, Singapore is apportioning 22% of its technology budget to this field alone.

Cyber-security has been described as integral to fortifying Singapore’s position as an international hub for business. Key to preserving Singapore’s innovations and ideas would be then to make sure it’s free from hackers and cyber looters. On a more physical level, in the wake of global atrocities, cyber-security could also forewarn Singapore of potential maliciousness directed at us.

The key to this lies again in data. One application of data that would then be effective here is data mining. Data mining is described as the finding of unknown properties of your raw data sets. The way this can be done is, again, through Machine Learning. The difference here then is instead of predicting an outcome, ML algorithms would be constructed to spot anomalies in the data set.

Crimes today take on many forms and is somewhat amorphous; one could then with access to large data sets and ML algorithms, identify outlying properties of data that was previously unknown. These, in turn, could help predict potential crimes. As algorithms help expedite processes, machine learning can be seen as tying with the government’s hopes for improving “the public sector’s response time to cyber security threats”. Furthermore, in view that threats to cyberspace have become more automatic in nature, it would be commensurate to counter AI threats with AIs of our own.

Conclusion:

I hope the knowledge I have distilled from my stint in Hackwagon thus far has been put forth succinctly and helped you understand, in the simplest possible way, why Singapore requires ICT professionals and more importantly, what skills are in hot demand.

In the course of my explanations, I also hope that you, the reader are motivated, just as I am now, to learn programming and better appreciate the beauty that is data science.