Challenges and Opportunities of the Asigmo Program: Part Two of an Interview with Iago Doel-Perez





Iago Doel-Perez, an enthusiastic researcher with a multidisciplinary background, talks to us about the challenges and opportunities he faced during his participation in the Asigmo Data Science Program.

Read all about Iago’s challenges and opportunities he faced while working on his PhD in part one of this interview.




Asigmo:

What exactly drove you to look at data science programs and challenge yourself to learn new things?

Iago:

Once I got a self-introduction into data science I first wanted to know, in comparison to other people, what are my big knowledge gaps, because obviously when you train yourself, you go through a certain path, but you often miss many elements that you are not aware of. That's what training does. There's somebody who can tell you that. Then there were all these new tools about deep learning that I was totally not familiar with, but I knew I wanted to reach. I saw this Asigmo Program that was perfectly fitting. It just gave me the sort of quality control to see where I am and certainly gave me a lot of new tools and new approaches on how to do this. Also, I have to say the extra training in agile was also very, very eye-opening. Likewise, filling other of those gaps that I would not have looked for myself, such as Hive & Spark, Natural Language Processing, cloud computing, End-to-end Machine Learning and Deployment or AI Ethics was fundamental.


Asigmo:

Am I understanding it correctly that you went into the program expecting to learn a lot of new tools, but also to expand the knowledge on the tools you already have, but then ended up learning even more with agile methods, for example?

Iago:

Yes, on one side it was to check whether I was on the right path. The other side was learning about these new tools especially in artificial intelligence or deep learning. And finally, I expected to learn things I didn’t know existed, and this was met. Just like the meme: I was not expecting certain topics, but I was expecting not to expect them, so I guess it doesn´t count as a surprise.


I knew that I didn't have the knowledge to decide what was what I needed. I put myself in the hands of professionals to tell me what is out there and what could I need that I am not even aware of.


Asigmo:

How did you feel about the interaction with the mentors? Do you feel like you were well-led, and you got good feedback during the whole program?

Iago:

Absolutely. So, I think every mentor was pretty different, but they all were very good. I really learned a lot and felt motivated all the time to learn more. The problem is sometimes it's only one week. So even if you want to stay longer, you can't. There is a, I would not say overlap, but there's a kind of redundancy, which I think was good. I think this happened with one of the professors that were very good, but for me, it was pretty hard to follow. These parts were also in a way covered by other mentors. I think in the end it was a very good balance. Also, due to the different backgrounds of the students, some topics move at a different speed. Maybe it's slower for you and faster for the other ones. But my impression actually was that it was very well-calibrated. I never felt bored at all. And I also didn't feel overwhelmed.


For example, during the first week, we had SQL packed in two days and it was quite intense. I thought to myself, if this is the rhythm of the rest of the course, I cannot make it. But then it balanced itself out in the following weeks. It was great. The progress in that aspect was amazing. There is also a lot of sandboxing, one-on-one with the mentors. It was also where the support came not only from the mentors but also the colleagues. So, if you can’t follow some topic, you can talk to your colleague and you try to help each other to teach each other what you just learned. And it is a very, very good way to learn, trying to discuss it with somebody, who is also new to it and with a different background.


Asigmo:

Talking about high intensity and also challenges, what was the most challenging part for you during the program itself? Where do you think you learned the most?

Iago:


For me coming into artificial intelligence was a start from zero so I learned a lot. For deep learning, we had one specific week on deep learning and then the natural language processing that took a lot from that. I guess that's where I would say, quantitatively, that I learned the most. Also learning different languages that I was not familiar with, like R and SQL.

As I mentioned before, SQL was challenging. In the beginning, I had the feeling of jumping into the cold water, but I don't think there was a point where I was overwhelmed. I think also the course adapts in a way to what you can expect. There are always exercises of different difficulty, and maybe you cannot make the top-level ones by yourself, but you can always learn from every lecture and every exercise, getting the important message.


Asigmo:

Was there anything that stood out to you, maybe a person or a field that you didn't quite expect to be in the course, some new perspective onto data science?

Iago:

For example, natural language processing. To be honest I didn't even know what it was when I first read about it in the syllabus. I googled NLP and I got neuro-linguistic programming, which was something about psychology. I thought, okay, this doesn't really fit in here. Then of course, with a bit more research I got that it is natural language processing and it was a big surprise. I really enjoyed it. I saw something that was very new and that I was not looking at but that it is extremely powerful and doable. Anyone can work with it, can do something out of it. So, it’s not just a showcase. It really is something that you can put to work.


Asigmo:

Who would you recommend this program to?

Iago:

The only demand is one should know a bit of Python or at least of statistics. I would recommend it to anyone. I would recommend it also to the people who are making the decisions up the chain. You don't have to become an expert, but you at least should be familiarized with all these terms and what is out there. I realized that at times the people that make the decisions have no clue about this kind of field. They need to know because this thing is growing so fast. Sometimes they are the bottleneck. I see that sometimes people have the wrong impression and they think that data science is going to solve all the problems and what people would need is a more educated vision on the field.

Knowing about the limitations, knowing what you can expect from the data and how to look for and what you could be biased too, what could lead to misinterpretation, and so on.

9 Ansichten0 Kommentare

Aktuelle Beiträge

Alle ansehen