AI Learning & Career Paths: Q&A with Dr. Samah Hijazi | Coursera Community
Coursera Header
Q&A

AI Learning & Career Paths: Q&A with Dr. Samah Hijazi

  • 7 January 2020
  • 52 replies
  • 4011 views
AI Learning & Career Paths: Q&A with Dr. Samah Hijazi
Userlevel 7
  • Community Manager
  • 1581 replies

Can you teach yourself AI? How should you be thinking about your AI career path?

If you’re newer to AI and have some questions about next steps, this is the Q&A for you! Data science researcher Dr. Samah Hijazi invites your questions on: 

  • starting your academic learning path in AI and Data Science,
  • learning AI and Data Science through MOOCs, and
  • building a strong portfolio for your career path in AI.

You’re also welcome to ask her any general Data Science questions.

Post your questions below and Dr. Hijazi will answer as many questions as she can from 15–30 January 2020.

About the Q&A Host

Samah Hijazi is a data science researcher who received her PhD in Machine Learning jointly from the Lebanese University and Université du Littoral Côte d’Opale in France. Her research interests include Machine Learning, Pattern Recognition, and Data Mining. In these areas, she is focusing on feature selection as a tool for dimensionality reduction and is passionate about making sense of such high-dimensional data since its visualization often requires a space of way more than the three dimensions we live in and perceive. Samah believes that we are living in an era of fast research advancements in artificial intelligence, so she aims at building algorithms that can uncover insights, extract knowledge, and bring value to humans.
 

Thank you to AI community leader @Mo Rebaie for organizing this Q&A!


52 replies

Userlevel 1
Badge +1

Hi. I need to build a staff recommendation system based on a bunch of staff features such as pay rate, grade, previous experience, types of training, past feedback by clients etc. I am just starting out in AI and the examples I have come across are very simple comparatively. I don’t know how to move forward with handling so many features. Any suggestions? Thanks.

Userlevel 1
Badge +1

Hi. I need to build a staff recommendation system based on a bunch of staff features such as pay rate, grade, previous experience, types of training, past feedback by clients etc. I am just starting out in AI and the examples I have come across are very simple comparatively. I don’t know how to move forward with handling so many features. Any suggestions? Thanks.

To add to the complexity the system has to work with the best possible staff matches to a given criteria of staff requirement; as exact matches between staff available and staff required are often not possible.

 

Userlevel 2
Badge +1

Hi Samah Hijazi,

Could you please guide me on what are the steps I need to follow to build my career in AI?

Userlevel 6
Badge +6

Cordial greetings,


What a privilege to continue knowing and learning from such talented people within this community. Thanks @Mo Rebaie for inviting this great Dr. as Miss Hijazi is, so that through her words we can go to understanding more, as well for those who want to start this journey.

Wanting to be part of this Q&A, could give us your opinion about what should be the mentality that a person must have to incur in this field, that is, if we take you as an example, how was your preparation or motivation for this career path.
In many cases we want to start something, but then we see it, and we perceive so many things, that we don't know where to start or what we should learn first, tell us in a comfortable way your beginnings in this science, also what you would recommend to women's for feel attracted and project their future in Artificial Intelligence, making reference to this post made by @Laura about "Artificial intelligence and the gender gap – what can be done?".

 

Thanks for the attention and my best welcome to you in this amazing community.

 

 

Regards.

Userlevel 4
Badge

Hi. I need to build a staff recommendation system based on a bunch of staff features such as pay rate, grade, previous experience, types of training, past feedback by clients etc. I am just starting out in AI and the examples I have come across are very simple comparatively. I don’t know how to move forward with handling so many features. Any suggestions? Thanks.

Indeed, not all the features describing the staff are equally important, however, you can only decide what features best fit your problem by specifying your objective function (e.g. hiring someone with the highest qualifications to do a specific job). A good example for you to start from is this paper: “Recommendations to Support Staffing Decisions” by Abigail Gertner, Susan Lubar, and Beth Lavender. Their goal is to match open job requisitions with appropriate staff.

Badge

Hello everyone,

Thank you @Laura and thanks to @Mo Rebaie for motivating us to learn at Coursera and improve our knowledge and skills. I have two questions for @SamahHijazi :

1- I graduated last year from the University with a Bachelor's Degree in Computer Communication, would you advise me to continue my academic path to achieve a Masters's degree in the University, or complete some online courses in Machine Learning at Coursera and start my career in Data Science?

2- What are some useful resources for building a strong portfolio in my career path in Data Science (books, resources,..)?

Thank you.

Userlevel 4
Badge

Hi Samah Hijazi,

Could you please guide me on what are the steps I need to follow to build my career in AI?

Hi Krish,

Well, to make it easier, we can start by categorizing AI career paths into two main groups: 

  • AI tool-user, where you only need to understand the available tools, algorithms, models, processes, and of course your data. In this case, you can boost any job you are already in by making it data-driven. The available tools to extract knowledge and uncover insights have become user-friendly with drag-drop capabilities (even without needing to code), so with some effort, it can be very easy to understand, learn, and apply different algorithms on your data. There is a large number of online courses that can put you on the right track to this path.

 

  • AI Algorithm-builders, where you need to have specific background knowledge in mathematics (probability, statistics, algebra, calculus, logic, and algorithms) and engineering. Algorithm builders are continuous learners that focus on the problems and shortcomings of current AI solutions. So, this mainly includes academic paths and research and development jobs.

Unlike many people think, AI is not only about being able to code but about analytical thinking and being able to analyze real-life patterns and transforming them into data solutions and models. For example, many optimization algorithms (a very important part of AI) like Annealing, Particle Swarm optimization, and Genetic Algorithm were inspired by aspects of nature. For example, Ant Colony Optimisation (ACO) is a population-based method inspired by the ant’s capability of finding the shortest path from the nest to a food source.

So, in which category would you like to start your AI career?

Userlevel 7
Badge +6

Hello, 

We are glad to meet you @SamahHijazi at our AI community!

Many junior data scientists always ask me a common question, and I'm excited to hear from you.

A lot of startups and even large AI companies are currently opening new job opportunities for junior data scientists/ junior machine learning engineers, does that mean to accept any job offer related to the field to start with, or wait till receiving an opportunity that better suits the domain of knowledge and academic background? 

Badge

Hi,

I’m a newbie in AI, done a few courses in coursera in AI. Through my time i spent in AI, i’ve come to love the CNN or more accurately AI image processing. Could you guide me through what should i do to make myself more close to the field? where should i start?

Badge

Hi Samah Hijazi,

Could you please guide me on what are the steps I need to follow to build my career in AI?

Hi Krish,

Well, to make it easier, we can start by categorizing AI career paths into two main groups: 

  • AI tool-user, where you only need to understand the available tools, algorithms, models, processes, and of course your data. In this case, you can boost any job you are already in by making it data-driven. The available tools to extract knowledge and uncover insights have become user-friendly with drag-drop capabilities (even without needing to code), so with some effort, it can be very easy to understand, learn, and apply different algorithms on your data. There is a large number of online courses that can put you on the right track to this path.

 

  • AI Algorithm-builders, where you need to have specific background knowledge in mathematics (probability, statistics, algebra, calculus, logic, and algorithms) and engineering. Algorithm builders are continuous learners that focus on the problems and shortcomings of current AI solutions. So, this mainly includes academic paths and research and development jobs.

Unlike many people think, AI is not only about being able to code but about analytical thinking and being able to analyze real-life patterns and transforming them into data solutions and models. For example, many optimization algorithms (a very important part of AI) like Annealing, Particle Swarm optimization, and Genetic Algorithm were inspired by aspects of nature. For example, Ant Colony Optimisation (ACO) is a population-based method inspired by the ant’s capability of finding the shortest path from the nest to a food source.

So, in which category would you like to start your AI career?

 

Hi Samah,

That was an incredibly good answer, thank you so much for your time.  Sorry @krish for interrupting your question.   I am absolutely interested in AI Algorithm-Builder carrier, I got a degree in Computer Science and I love Math.  So I would like to improve my Math knowledge for that particular area.  I have been thinking to take a Math Degree but I am not too sure.  What do you recommend?  Is there any good online Master Degree to become AI Algorithm-builder?

Thank yo much!

Badge

Hi Samah Hijazi,

Could you please guide me on what are the steps I need to follow to build my career in AI?

think of problems to be solved by AI and gain skills in AI and then apply them to solve the problems

Userlevel 1
Badge

Dr. Hijazi,

Thank You for taking questions. I am really excited about Python and Data Science. I am retired from International Medical Robotics Applications. I have a BS and MBA. I have been taking MOOC’s on Python and Statistics for 4 years.

I wonder, is there room for a 65 year young guy in the Data Science field, or is it pretty much a young persons only field??? I ask because I am not sure if I want to invest $40,000 in tuition just for information. I’d like to be able to use new found knowledge.

Thank You,

-Bill Giles

Badge

Hello Dr.Samah,

 

I’m doing a research in defining the uncertainty in deep learning models as a part of my master’s thesis. Either uncertainty in the model itself or the uncertainty of the data.

Actually I find it very important topic which concerns the safety of the machine learning applications yet it’s ignored by machine learning researchers, do you agree with me?

And can you recommend me resources or materials that can help me with my research ?

And any research points in this topic you can suggest, that I can use it for my thesis?

Thank you, 

Badge

Hello Dr.Samah,

I’m really interested by the field of AI, I spend a lot of time learning about it, did most online certifications , implement things using Keras or Pytorch and read academic papers, I already have a master degree in physics but it seems that I cannot transition, companies ask for a Phd in AI or more.

Should I go back to university and start over from zero or do you think there is a path to transition at least in the industry ?

Badge +1

Hello Dr. Samah, 

I am an experienced professional in the field of wireless and telecommunications. I am 54 yrs old and it is becoming difficult to get jobs or projects, so looking forward to venturing in new technologies. I believe that knowing AI/ML/DL will help me in Engineering areas,  and other areas, including business & finance. 

I have completed one course in Machine Learning from Coursera and did a course in Data Analysis from Udacity. I am also planning to do the Deep Learning specialization consisting of 5-6 courses. I am expert in Matlab and Python. 

Meanwhile I was trying for jobs (in India) but there is absolutely no response although there are plenty of positions. So I was baffled and had doubts. What bothered me was the unknowns - was  it my age or lack of actual openings.

So I need your inputs.  

  1. Will doing more certifications from Coursera help me in the alternative career in terms of finding highly paid jobs?
  2. How do you see the field with respect to budgets, future potential and growth? 

Pl let me know frankly what you think. 

Best, 

Jaydeep Roy 

Bangalore, India 

Badge

There is increasing need for managers, who have technical skills in analytics also. How data scientist can develop soft skills?

Userlevel 4
Badge

Hello everyone,

Thank you @Laura and thanks to @Mo Rebaie for motivating us to learn at Coursera and improve our knowledge and skills. I have two questions for @SamahHijazi :

1- I graduated last year from the University with a Bachelor's Degree in Computer Communication, would you advise me to continue my academic path to achieve a Masters's degree in the University, or complete some online courses in Machine Learning at Coursera and start my career in Data Science?

2- What are some useful resources for building a strong portfolio in my career path in Data Science (books, resources,..)?

Thank you.

 

Hi Ali, 

Regarding your first question, I believe that since you already hold a degree in a computer-related field, you do not really need to enroll in a Master's program in order to work in the AI domain. In my opinion, university courses will not turn you into an expert, these will introduce you to the basics. You will always need to put personal effort and take a self- learning track in AI. So, I think that in your case online courses are a good idea.

Keep in mind that if you are thinking about having a Master's degree and then going for a Ph.D., this is a personal preference. For example, I took an academic path because it was my passion. So only you can decide on this.

here is a book that I think can be a good start for you:https://repo.palkeo.com/algo/information-retrieval/Data%20mining%20and%20analysis.pdf 

Userlevel 4
Badge

Hello, 

We are glad to meet you @SamahHijazi at our AI community!

Many junior data scientists always ask me a common question, and I'm excited to hear from you.

A lot of startups and even large AI companies are currently opening new job opportunities for junior data scientists/ junior machine learning engineers, does that mean to accept any job offer related to the field to start with, or wait till receiving an opportunity that better suits the domain of knowledge and academic background? 

Hello Mo, 

Thanks for your kind invitation to this great community, I am glad to be here.

The best thing about data is that the more you apply the more you learn. So, taking available opportunities regardless of the position title will always be an added value to one’s learning curve, experience, and skills. Also, sometimes the perfect fit can be the junior position itself.

Userlevel 4
Badge

Hi Samah Hijazi,

Could you please guide me on what are the steps I need to follow to build my career in AI?

Hi Krish,

Well, to make it easier, we can start by categorizing AI career paths into two main groups: 

  • AI tool-user, where you only need to understand the available tools, algorithms, models, processes, and of course your data. In this case, you can boost any job you are already in by making it data-driven. The available tools to extract knowledge and uncover insights have become user-friendly with drag-drop capabilities (even without needing to code), so with some effort, it can be very easy to understand, learn, and apply different algorithms on your data. There is a large number of online courses that can put you on the right track to this path.

 

  • AI Algorithm-builders, where you need to have specific background knowledge in mathematics (probability, statistics, algebra, calculus, logic, and algorithms) and engineering. Algorithm builders are continuous learners that focus on the problems and shortcomings of current AI solutions. So, this mainly includes academic paths and research and development jobs.

Unlike many people think, AI is not only about being able to code but about analytical thinking and being able to analyze real-life patterns and transforming them into data solutions and models. For example, many optimization algorithms (a very important part of AI) like Annealing, Particle Swarm optimization, and Genetic Algorithm were inspired by aspects of nature. For example, Ant Colony Optimisation (ACO) is a population-based method inspired by the ant’s capability of finding the shortest path from the nest to a food source.

So, in which category would you like to start your AI career?

 

Hi Samah,

That was an incredibly good answer, thank you so much for your time.  Sorry @krish for interrupting your question.   I am absolutely interested in AI Algorithm-Builder carrier, I got a degree in Computer Science and I love Math.  So I would like to improve my Math knowledge for that particular area.  I have been thinking to take a Math Degree but I am not too sure.  What do you recommend?  Is there any good online Master Degree to become AI Algorithm-builder?

Thank yo much!

Hello Luis,

you are most welcome. I am very happy to see the excitement between your lines.

About the AI Algorithm-Builder career, I do not think you can find straightforward courses to teach you that.

It is an exciting self-learning process. After understanding the basics (you should already understand algorithms, data types, learning contexts, the required maths, etc.) as much as you can, you move towards reading new research articles. Of course, when you start reading you will focus on one subject that you find interesting. It is a very good idea if you even try to replicate the work done to see whether you got really right (it won't be easy). Eventually, you will find yourself able to suggest amendments to the algorithms at hand in order to tackle a specific problem that you have encountered. You finally start building prototypes.

Also, yes you are right about improving your mathematical background. However, you do not have to go for a math degree only for this reason. For machine learning, you should focus on specific mathematical topics like Linear Algebra.

Here is a 4-minutes-read that I found concise in explaining mathematical needs: https://towardsdatascience.com/the-mathematics-of-machine-learning-894f046c568#--responses

In my opinion, you can go for a Ph.D. degree in AI, it will take you approximately through the same path.

I wish you the best.

Userlevel 4
Badge

There is increasing need for managers, who have technical skills in analytics also. How data scientist can develop soft skills?

Hello Anna.Korolyuk,

Well, there is a lot of trainings on interpersonal skills like presentation skills, communication skills, time management, project coordination, project management and so on. These can be considered. Also, some on-job practices can be beneficial like stepping out of the technicalities to see the big picture and sharing ideas. For example, data scientists need to communicate with front-line domain experts because they are the ones who understand business needs the most. Collaboration and presenting their insights is a key.

 

Badge

Hi Samah,

 

What are you thoughts on a Msc in Data Analytics and Machine Learning graduate? In specific to the job market, are they disadvantaged against PhDs? Or is there a enough and differing roles within the industry for these two groups of grads. I am within the EU/UK region if that helps narrow down the scope for your answer. 

Thanks!
David

Userlevel 4
Badge

Hi,

I’m a newbie in AI, done a few courses in coursera in AI. Through my time i spent in AI, i’ve come to love the CNN or more accurately AI image processing. Could you guide me through what should i do to make myself more close to the field? where should i start?

Hello ridwanarfeen,

What courses have you taken? I mean whats the level of your background knowledge.

 

 

 

Userlevel 4
Badge

Dr. Hijazi,

Thank You for taking questions. I am really excited about Python and Data Science. I am retired from International Medical Robotics Applications. I have a BS and MBA. I have been taking MOOC’s on Python and Statistics for 4 years.

I wonder, is there room for a 65 year young guy in the Data Science field, or is it pretty much a young persons only field??? I ask because I am not sure if I want to invest $40,000 in tuition just for information. I’d like to be able to use new found knowledge.

Thank You,

-Bill Giles

Hello Bill,

Thank you for sharing your concerns. First of all, you have built a good background already, in my opinion, 4 years on python and statistics are enough for you to start working in the field. Regarding your age, it depends on the culture of the company you are willing to work for. I do not think it is a barrier, especially if you make use of the experience you have gained during your years of work. With your educational background, experience, and technical knowledge (python and statistics) you can invest this amount in building your own startup in healthcare for example.

Best,

Userlevel 4
Badge

Hi Samah,

 

What are you thoughts on a Msc in Data Analytics and Machine Learning graduate? In specific to the job market, are they disadvantaged against PhDs? Or is there a enough and differing roles within the industry for these two groups of grads. I am within the EU/UK region if that helps narrow down the scope for your answer. 

Thanks!
David

Hi David,

Data science is a huge domain now with several job roles emerging across all industries. For example the emerging role of the so-called “Data Translator”. Few years back, different capacities were set to be required by one individual who should also be a PhD. However, the industry recognized that it is a burden for one individual to have all these skills. Indeed it is a team sport now where each person has a part to play. Thus, you might have a specific skill that can differentiate you in the data science job market regardless of the academic level. 

Badge

Hi David,

Data science is a huge domain now with several job roles emerging across all industries. For example the emerging role of the so-called “Data Translator”. Few years back, different capacities were set to be required by one individual who should also be a PhD. However, the industry recognized that it is a burden for one individual to have all these skills. Indeed it is a team sport now where each person has a part to play. Thus, you might have a specific skill that can differentiate you in the data science job market regardless of the academic level. 

 

Thanks for you input Samah! Appreciate you taking the time to share.

Interesting point about the ‘data translator’ role, hope to see this pick up within the EU/UK region. At this point, I think the PhD’s are suited for specialized tasks such as algorithm creation. I’ll work on developing other skills around this field then, as math is not my strong point! Cheers.

Reply