Getting started on AI

The PM has repeatedly called on Malaysia's education curriculum to be revamped to focus on new technologies such as Artificial Intelligence. In this article, the Rembau Times gives our views on how this can be accomplished.

3
587

The Honorable Prime Minister Tun Dr Mahathir had recently called on Malaysians to master skills in Artificial Intelligence. He mentioned this call over here in his speech to the Malaysian diaspora in Singapore on Nov 12.  More recently Education Minister,  in an interview with Malaysiakini on Dec 04, Maszlee Malik mentioned that the PM wanted an overhaul of what students are taught to include technical knowledge such as artificial intelligence and coding.

The PM has called on Malaysians to develop competencies in new technologies such as AI.

Well its currently December 09th and we have not heard anything else on this subject. Of course, the PM should not be the one to come out with policy and curriculum, that is why we have a Cabinet and a Ministry with thousands of staff to get the job done.

However, in the spirit of helping achieve the Prime Minister’s vision, let me share a bit about what can be done to kick start this initiative. One field AI can work is to help develop intelligent agricultural systems in Malaysia, where we can reduce the reliance on foreign labour as well as develop systems to be more effective in utilizing resources to grow different types of crops, including fruits and vegetables.

Palm Oil prices have slumped and looks very unlikely to recover as Indonesia and Africa will supplant Malaysia as the lowest cost producer. Malaysia needs to innovate.

One of the limiting factors to achieve this is the lack of teaching / lecturing talent in Malaysia. So, in a way the country is hampered because without a pool of teaching talent, it is very difficult to train the next generation workforce to understand these skills.

In this regard, there are actually some very well structured online learning courses on Artificial Intelligence. The one, which he happened to be enrolled in is by Coursera’s Machine Learning. To help everybody, Artificial Intelligence and Machine Learning roughly refer to the same thing.

Prof. Andrew Ng from Stanford University is an expert in the field of Machine Learning. He has worked previously with Google and Baidu on cutting edge projects.

The course is conducted by Stanford Professor Andrew Ng, who happens to be the founder of Coursera. The course fee is about USD$80 (yes about RM 350 only) and it runs for 11 weeks. All the content is delivered via online video lectures and it includes an assessment component as well. We have actually enrolled in it and are 2 weeks into the course so we can give some first hand feedback.

The syllabus is extremely well structured. It starts with giving an overview of what Machine Learning / AI is about. Week 2 teaches students to implement their first learning algorithm using the Gradient Descent technique to optimize a set of parameters against a hypotheses. Week 3 is on Logistic Regression, Week 4 – 5 on Neural Networks, Week 6 on applying machine learning, Week 7 on Support Vector Machines (this is a really powerful technique), Week 8 on Supervised Learning, Week 9 on Anomaly Detection and Recommender systems, Week 10 on Large Scale Machine learning and Week 11 on developing an image recognition application. The syllabus follows the Stanford Introduction to Machine Learning course, so its very high quality.

The key prerequisite is that students must understand English and have mastery over linear algebra (i.e. matrix operations), some experience in programming, as well understand calculus. All of this should be covered by first year university in any technical discipline – Electronic & Electrical Engineering, Computer Engineering, etc. It is very well suited in year 2 course in our local universities.

However, another prerequisite is the honour system. In line with the Government’s decision to stamp out corruption, there should also be no cheating or plagiarism. The assessments are automarked and the results are available immediately.

We call on Minister Mazlee to liaise with respective University Vice-Chancellors in order to quickly get started on Malaysia’s AI initiative for the benefit of the country.

 

 

 

3 COMMENTS

  1. until the best student is allowed to take up courses they applied for and not denied a place in uni because of his/her race or religion, there is no hope. you can educate a B student to be the next Einstein. Just look at Proton, if only they hired the best of the best to for the CEO and Engineering team, and no, I am not saying Chinese Engineers are smarter than Malay Engineers. What I am saying is Proton should have hired good engineers from Germany, USA or Japan…. but here, the criteria is race and religion.

    Until we do away with all these race based policies, nothing will change.

  2. Er i think the issue is ppl shd not give up. tdy any person who is upset abt not getting the chance to do computer science cld go on Github learn to code think of a great idea and get hired to do tech.

  3. Just finished week 3. Used a logistic regression with regularization and an uncomstrained minimization algo to calibrate the parameter vector. Example was on a semiconductor defect analysis. Just learnt all of this in the last 2 days. This is machine learning course is defn top rated but understanding of matrix and vector operations is a must.

LEAVE A REPLY