Q&A with Kunyu, IBM Data Science Professional Certificate alum | Coursera Community
Coursera Header
Q&A

Q&A with Kunyu, IBM Data Science Professional Certificate alum

  • 30 August 2019
  • 10 replies
  • 1151 views
Q&A with Kunyu, IBM Data Science Professional Certificate alum
Userlevel 7
Badge +9
If you're currently working through the IBM Data Science Professional Certificate and need a little inspiration to keep going, or if you've been thinking about taking the certificate, now's your chance to talk to someone who has completed it.

Kunyu recently completed the certificate and has since managed to get an internship. Between September 2nd and 8th he'll be answering any questions you have for him about his experience. Post your questions below and he'll answer as many as he can over the coming week.

About Kunyu
"I'm a second-year master's student in Computational Analysis and Public Policy at The University of Chicago. I brushed up my skills in Python programming, acquired comprehensive knowledge of applied machine learning, and practiced hands-on in the IBM Cloud in the IBM Data Science Professional Certificate program. With the job-ready skills and relevant project experiences, I got the opportunity to serve Lasalle Investment Management on the Research & Strategy team as a Data Science intern in Chicago for summer, 2019."

10 replies

Userlevel 1
Badge
Congratulations Kunyu.
It would be interesting that other Alumni share their experiences : how this course helped them in their careers and / or get a now job.

Currently I work in the Medical Device Industry, where Data Science and Machine Learning are used more and more often, mainly in the Imaging area.

I think, such as Kunyu, Finance is probably a major place to work in for with Data Science Background.
Badge
Hello Kunyu,
Greetings,
I am also completing the IBM Data Science Professional Certificate course. Me and other fellow members have faced too much confusion in the " Python for Data Science and AI " course.
I am sure you might have faced the same or some confusion in this course.
Please can you guide us how can we fulfill the deficiency and gap of this course's skills. Is there any other online platform which is good to learn from?
Thank you
Userlevel 1
Badge
Congratulations Kunyu.
It would be interesting that other Alumni share their experiences : how this course helped them in their careers and / or get a now job.

Currently I work in the Medical Device Industry, where Data Science and Machine Learning are used more and more often, mainly in the Imaging area.

I think, such as Kunyu, Finance is probably a major place to work in for with Data Science Background.


Hi AxeIG,

I didn't say or imply that I see finance as a major place to work in for people with data science background. I see it as a field that I would apply my skills to make business impact, along with Tech.

In terms of education, I have a background in Economics and International Relations back to my undergrad. Then I proceed to a interdisciplinary master's program in computer science and public policy.

In terms of profession, I interned as investment banking analyst twice during my undergrad, and I interned in data anlyst/science roles in tech and finance during my senior-year and my master's study.

To my understanding I do not have a data science background, but rather one in economics and social sciences on a quantitative track. It's hard to say that anyone comes from a data science background, since the undergrad students who major in DS or AI mostly haven't graduated yet. Especially for applied data scientist (research data scientists are mostly in math or computer science), it's a combination of your background and data science skills.

I would say I have data science skills, Python, R, Data Cleaning, Data Viz, Machine Learning... and given my background and skills, the domains I chose are finance and tech.

For applied data scientist, domain knowledge is of critical importance. Medical and healthcare is a great area for people with a background in medicine. Skillwise, for time series analysis and image recognition, I would say deep learning could be a of great value.

Getting a job whether in finance, tech or healthcare as an applied data scientist, I think it's important to show you have both skills and domain knowledge, or at least enough interest to understand and work alongside a seasoned professional in the field you choose.

Kunyu
Userlevel 1
Badge
Hello Kunyu,
Greetings,
I am also completing the IBM Data Science Professional Certificate course. Me and other fellow members have faced too much confusion in the " Python for Data Science and AI " course.
I am sure you might have faced the same or some confusion in this course.
Please can you guide us how can we fulfill the deficiency and gap of this course's skills. Is there any other online platform which is good to learn from?
Thank you


Hi Jehangir,

Great question! I think the IBM Data Science Professional Certificate Program gives you a comprehensive introuction to the big picture, but many details that the instructors refer to as "beyond the scope of this course" could be gaps you are referring to.

For example, I think Python programming skills, whether it's algorithms or more applied ones like with Pandas, NumPy and Matplotlib can be a place to start. For people without prior exposure or systematic education to it, it's really hard sometimes to find the right answer for the problem they tackle. However, mostly it's the case that they are used to copy and paste off-the-shelve codes from others and wouldn't like to know about how it works. When it comes to cases where the dataset is large enough and their Pandas code is taking forever, or they're asked to add a feature to your quite standard Matplotlib plot, people would eventually see the gap.

In fact, diving deeper into Python programming definitely worth the efforts. In the cases I mentioned above, learning about time complexity and Matplotlib Axes goes a long way. You might still have to Google a lot, but you would know which snippet to use and how to customize it for your needs.

The other gaps really depend on your need: For machine learning you might want more math and understand why the algorithm works in that way sepcifically and why changing the hyperparameters would make a difference in a specific way; For databases you might want to learn more about SQL and relational databases, NoSQL and Hadoop; For cloud computing you might want to learn Spark, and AWS/Azure/Google Cloud‎.

Other platforms I would recommend... I see library documentations as a great source, like for scikit-learn and NumPy. Also Kaggle can be one. For me it's also great to learn from books/e-books, especially for Python programming and algorithms.

Kunyu
Badge
Thank you Kunyu .......... Great answer 🙂 ......
I have registered with kaggle. It looks very useful at the moment.
Can you please explain library documentations in a more simple way. I could not get you here.
And I saw that kaggle is also offering only introductory level Python course lessons.
Can you please guide me towards a website where I can master Python and algorithms step by step in a easy and comfortable way.
I will really appreciate your help.
Kind Regards,
Jehangir
Userlevel 4
Badge +4
Hello Kunyu,

you are doing a great work, I wish you best of luck!

I have two questions for you:

1- What are your thoughts about the rapid rise of federated learning that could take place in the near future?

2- Do you think that the IBM professional certificate specialization is a solid start in Data Science for beginners?

Thank you.
Userlevel 1
Badge
Thank you Kunyu .......... Great answer 🙂 ......
I have registered with kaggle. It looks very useful at the moment.
Can you please explain library documentations in a more simple way. I could not get you here.
And I saw that kaggle is also offering only introductory level Python course lessons.
Can you please guide me towards a website where I can master Python and algorithms step by step in a easy and comfortable way.
I will really appreciate your help.
Kind Regards,
Jehangir


Hi Jehangir,

I think Udemy could be a nice source for introductory courses. From my experience courses there are less structured but can be more specific in terms of topic (like Python Data Structure For Interviews) and might suit your needs. This can be good and bad though. One very personal point that I would like to make is do not use DataCamp, if you're serious about learning things in depth.

Here is a link to scikit-learn documentation, you can come to the contents when you're confused about how to fit and tune a machine leanring model.

Kunyu
Userlevel 1
Badge
Hello Kunyu,

you are doing a great work, I wish you best of luck!

I have two questions for you:

1- What are your thoughts about the rapid rise of federated learning that could take place in the near future?

2- Do you think that the IBM professional certificate specialization is a solid start in Data Science for beginners?

Thank you.


Hello Mo Rebaie,

Very good questions. Personally I don't know about federated learning until this very moment. I worked for a HPC center on campus part-time and get to know parallel computation on multiple compute nodes with Slurm. I kind of understand the motivation of the new approach because the resources are quite expensive whether at my center or in the cloud. But to my understanding certain models cannot be trained online, and most of the use cases are still supervised. I wouldn't comment more on this because I'm really not familiar with the topic.

To your second question, I would give a cautious yes. The IBM Data Science Professional Certificate program has very well-designed curriculum that gives a big picture, yet the details requires more work. For beginners I think it's a great introductory course, but without further learning I wouldn't say it makes you a solid applied data science professional.

Kunyu
Userlevel 4
Badge +4

Hello Kunyu,

you are doing a great work, I wish you best of luck!

I have two questions for you:

1- What are your thoughts about the rapid rise of federated learning that could take place in the near future?

2- Do you think that the IBM professional certificate specialization is a solid start in Data Science for beginners?

Thank you.
Hello Mo Rebaie,

Very good questions. Personally I don't know about federated learning until this very moment. I worked for a HPC center on campus part-time and get to know parallel computation on multiple compute nodes with Slurm. I kind of understand the motivation of the new approach because the resources are quite expensive whether at my center or in the cloud. But to my understanding certain models cannot be trained online, and most of the use cases are still supervised. I wouldn't comment more on this because I'm really not familiar with the topic.

To your second question, I would give a cautious yes. The IBM Data Science Professional Certificate program has very well-designed curriculum that gives a big picture, yet the details requires more work. For beginners I think it's a great introductory course, but without further learning I wouldn't say it makes you a solid applied data science professional.

Kunyu


Thank you for sharing your thoughts!
Userlevel 7
Badge +9
Huge thanks to @Kunyu He for taking the time to answer questions about his experiences! The time window for getting answers from Kunyu has now passed but, @Kunyu He, I hope you'll continue to participate in this and the alumni community!

Also thanks to @AxelG @Jehangir @Mo Rebaie for asking thoughtful questions.

We will continue to invite learners to share their stories with the community and take some time to answer your questions so stay tuned!

Reply