Introductions thread - please say hi!

Hi everyone - please post here and let us know a little about who you are and why you’re here… I’d love to hear where you are joining us from, as well as:

  • How did you find out about the course? Have you used your knowledge from part 1 on any fun projects yet?
  • What are you hoping to do with deep learning?
  • What are you passionate about?
  • What’s something that not many people know about you?..

Here’s my answers:

  • I’m originally from Melbourne, Australia, which is a really nice town, but also rather boring… now I’m in San Francisco, which is never boring!
  • I learnt a lot by developing the material for part 1, and even more for part 2! I’m particularly building on my learnings from the NLP and embeddings material in creating sequence to sequence with attention models for neural translation - we’ll be learning about that soon…
  • The Kaggle data science bowl on lung cancer diagnostics has inspired me to look at this problem again and see if I can improve on the results I got a couple of years ago
  • I’m passionate about good food (I particularly love the food in Australia, Malaysia, and Japan), and (of course) about using data to solve impactful problems
  • I’m somewhat obsessed by all kinds of gadgets - especially “last mile vehicles”; I have 4 pairs of roller blades, 2 bikes, 3 electric skateboards, and electric scooter, and regular scooter, and more!..
1 Like

Hello everyone! My name is Brad, I’m a grad student in USF’s MSAN program and an intern at with Jeremy and Rachel. I was in Part 1 of this course, and worked on curating the course notes for the MOOC so if you have any questions feel free to reach out to me!

  • I’m very excited for Part 2 of this course. In Part 1 we learned a lot about CNN’s for image recognition (and more) as well as RNN’s for NLP, and I’m really interested in learning more about the exotic architectures we covered in Lesson 7.

  • I’m really interested in seeing how far Transfer Learning can take us in solving new problems. Given that most of us likely do not have access to terabytes of data to build models on (nor the resources for that matter), I’m excited to find out exactly how far we can take pre-trained models in solving similar tasks.

  • My undergraduate background is in mathematics, and math is something I’m very passionate about. I’m slowly learning more about computer science in my graduate studies, and I’m very interested in continuing that. I’m a firm believer in always understanding what’s under the hood. I’m also very passionate about music.

  • Some interesting things about me: I’m a big fan of trying new and different food. I’m also really into heavy metal, jazz, and progressive music (and any fusions of the three).


I’m Samar and I’m a Research Assistant at the Center for Language Engineering in Lahore, Pakistan, where I work on using machine learning and natural language processing to improve support for low-resource languages like Urdu, Pakistan’s lingua franca.

I had the opportunity to be a part of Deep Learning Part I as an International Fellow back in October. Participating in the course remotely was an incredibly valuable experience and taught me so much about the practical aspects of model design, implementation, and evaluation. Jeremy, your lectures were always peppered with useful tips and tricks on getting deep learning to work, and discussions on the Slack channel and course forum were a welcome escape from struggling with deep learning in isolation.

You might be pleased to hear that your course has already helped me make advances in my work on low-resource languages.

One of the biggest challenges I had witnessed during my research was the lack of labeled data available for training machine learning algorithms, a phenomenon that commonly hinders research on languages with little data on them. This got me thinking of ways in which unsupervised learning techniques can be employed to extract meaningful representations for use in supervised learning problems later on.

Taking a cue from Lesson 5, I acquired, cleaned, and segmented into sentences an Urdu corpus with over 35 million tokens and trained a continuous bag-of-words model to learn vector representations of words from it. The resulting embeddings captured not only very useful semantic relationships between words but also lexical variations frequently found in Urdu. This marks the first time such word representations have been trained for Urdu, and, while they are themselves an incredibly valuable resource, it is exciting to think of ways in which they can be used to advance the state of natural language processing for Urdu in applications ranging from text classification to sentiment analysis to machine translation.

Applications of recurrent neural networks demonstrated towards the end of the course have inspired me to test their effectiveness at building character-level language models with long-term dependencies for Urdu. I look forward to seeing how well they can capture the rich morphology that Urdu exhibits.

In the long run, I hope to use deep learning techniques to bridge gaps in human communication by helping computers better process and understand regional languages and use machine translation to help unlock a world of information for people who don’t speak English (or other popular languages).

I’m passionate about learning new things and sharing that knowledge with others. I have volunteered as a TA multiple times during my undergrad and I love explaining things by deconstructing complex equations into intuitive concepts. I hope to one day become a professor and get to do this full-time.

Something not many people know about me? That I was homeschooled right up to grade 10! That’s when I fell in love with learning and reading.

I can’t wait for the course to begin!

- Samar


Hi everyone, Even here. I’m a data scientist from Vancouver, Canada and used to work for a major online dating site. I’ve left there to retrain and refocus my career and coming up to speed on deep learning is currently my full time focus with the exception of a little consulting.

I’m passionate about art, and I’d love to be able to work on some generative neural nets along the lines of deepdream. I’m also really enjoying this time off of work with my 19 month old son. The course has me thinking about ways to generate childrens books via deep learning and we’ll see where that project goes. Like @jeremy I’m also passionate about food so maybe we can come up with a recipe/cookbook creation algorithm assuming we can find the right dataset.

I’m hoping to learn a number of different ways to apply deep learning to real world problems. After seeing the power of deep learning in Part I i’m hooked.

Here’s my linkdin profile and I’d love to connect with other students there as well to network and share opportunities/endorsements there.


Hi, Constantin here. I am joining from Germany. I live near Heidelberg.
By trade I am a product manager at a major optics manufacturer, hence I am interested in applications regarding biomedical image data.
I am currently taking time off work to care for my two children 1 and 5 years old. And I am remodeling myself with learning about data science methodologies with my major focus now being Deep Learning.
I am excited to hear you, @jeremy, are looking into the lung cancer problem again. I have spent considerable amounts (too much) time wrangling (almost wrestling with) the data, but I have come to terms with it using bcolz (great learning from course 1!).
Like @jeremy, @Even I love food (who doesn’t?) and do most of the cooking at home. Food is also a good way to get to know other cultures, which I had some opportunity to do during traveling. I love Italian, Thai, Indian and some Chinese food. Plus grandma’s best traditional recipes.
I also love music, both making and listening to, though the making part has been falling short during the last few years.
Another interest are recumbent bikes. I am still searching for the perfect recumbent with full fairing that is as practical as my car, but human powered (or electrified).

Hi Everyone!

My name is Radek and I am from Poland. I have only a year and a half of college (mostly sociology) and do not have formal math nor CS education.

I am a self-taught Ruby on Rails developer with what it seems to be a passion for machine learning :wink:

To answer @jeremy’s questions:

  • I stumbled across the course not too long ago on hackernews. Still in the process of completing part 1 but really looking forward to share in the adventure that taking part 2 is going to be!
  • What am I hoping to do with deep learning? I think that first and foremost I would like to satisfy personal curiosity and have some sense of achievement in this space. If I would have a chance to do deep learning / machine learning professionally, I think that would be great. Also, if at some point I could leverage this knowledge to contribute to helping others - be that through improving medical diagnostics or making it more accessible (like in the x-ray example that @jeremy mentioned in one of his videos) or via some other means, that would be extremely gratifying and would love to pursue it even should this be on my free time. Ah and one thing I would have forgotten - I am hoping to continue to have a lot of fun!
  • What am I passionate about? Learning
  • What’s something that not many people know about me? Last year I lost 20 kgs by picking up running! Gained a few of those back over the winter but hopefully will come back to running now that the winter is winding down.

I would also like to say that I continue to be blown away by the quality of the material that @jeremy and @rachel have put together. The videos, the wiki, the forums, the gradual introduction of concepts, notions that distill and convey deep insights about the field… this and way more that I am not even able to put into words is simply amazing.

Thank you very much for having me!


Hi everyone,

I’m Jonas. I did part 1 and it was great! I am glad to join part 2 as an “international fellow”.

I work as a senior research scientist at Siemens in a research field focused on artificial intelligence – the traditional approach (symbolic / model-based AI) as well as machine learning and in particular also deep learning. For my profile, please refer to LinkedIn at

Some more or less related courses I have been through that I can recommend for others:
Machine Learning - Ng (Coursera)
Statistical Learning by Hastie / Tibshirani (Stanford Online)
Mining Massive Datasets by Ullman (Stanford Online)
Berkeley AI course by Klein / Abbeel (edx).
I was actually a voluntary teaching assistant for this last course, UC BerkeleyX CS188 in Artificial Intelligence (MOOC) in 2015. Great course as well.

I want to use the knowledge I gain for my work, but this is also a personal passion of mine. I actually gave a 1-hour presentation to my colleagues in my group on deep learning, in large part based on what I learned from this course.

I very much like the format of the course, and I agree that much of deep learning material is not so accessible and I haven’t seen anything comparable to this approach. So I am very grateful, and much looking forward to the course.

Thank you Jeremy and Rachel!


1 Like

Hi Everyone, I am Renjith and I work as a Database Admin in Irving, Texas. I like to solve problems and gets easily addicted to anything that is challenging.

I don’t remember exactly what led me to Kaggle but when I went to the site I felt this might be the most interesting place to hang around. So I started reading, learning and practicing Data Science in my spare time. I even joined a Graduate Course for Data Science. I also realized the amazing potential for these skills to influence the space around us.

I want to apply Deep Learning to help the farmers of my home town in India. But my immediate goal is to work on the Kaggle Lung Cancer Detection problem and learn as much as possible along the way.

I like mild spicy food.

I like to live in the mountains ever since I last visited Broken Bow in Oklahoma. Something I might consider when retiring :slight_smile:



My name is Christopher, and I work in the information security industry and a little game development on the side.

A friend of mine mentioned the course to me. I was studying SciKit learn and was staying around from neural networks until I had learned Scikit learn well. I really wish I didn’t wait but very thankful I found this course. I’ve learned a ton in a very short time and super excited about part II.

I haven’t nailed down exactly what I want to do with machine learning, I have a few ideas but still up in the air. I want to do some of the harder Kaggle competitions like lung cancer once I feel a bit more confident and then work on a project of my own.

I love technology, gaming (video & board), photography, movies, maker (Arduino, Raspberry Pi, so on), automation. What I enjoy most is spending time with my son (9), usually gaming together and generally rough housing like brothers. iI’m also a huge Firefly fan.

Most people (except my wife) don’t know I have a difficult time not tearing up when someone I like dies valiantly at the end of a movie or at extreme frisson moments. I’m not exactly proud about it, but I’m not all that embarrassed either.


Hello all,

My name is Xinxin, I work for a small data consulting firm at Seattle. I built my first deep learning model with caffe, alexnet, and in my first fit I set 1000 epochs at 0.0001 learning rate. I had no idea what I was doing. Then I heard about fastai course on hacker news, and it’s been a dream adventure since.

I like all the subjects in part 2(GANs, reinforcement learning, attention model), and I am eager to discover how to do them in excel! I plan to apply deep learning to Parkinson’s disease patient care and fashion/ lifestyle. My journey to machine learning has been an empowering one. I begin to see myself as a technology creator (not just a consumer), and deep learning is the crucial piece in many technologies I am passionate about. I have a background in zero-emission technology, and I would love to see deep learning applying to energy technologies, although there doesn’t seem to be exciting applications or serious interest given the low oil price.

Something interesting: I love reading, I was really psyched after reading Peter Diamandis book Abundance and it’s amazing to follow the technology changes since 2012. My recent book crush is a type of German book called Wimmelbilderbuch, or a hidden picture book. My 3 year old and I can read these books together in a million different ways over and over again.

@Even, I like your children’s book idea. In fact, I was thinking about deep learning with Wimmelbilderbuch last night, it’s perfect for multi-object classification.

Here’s my linkedIn profile, let’s connect.


HI I am Vishnu Subramanian from India. I work for an Analytical company where I help couple of projects on Big Data particularly Apache Spark. And I am super excited about Deep learning and eagerly waiting for the course to start.

Answering Jeremy Questions

  1. I came to know about the course through an article on

  2. I am not exactly clear on what I am going to do with Deep Learning. But exploring a lot of things which include
    teaching , blogging , setting up a remote control to identify objects and probably learn to drive around my kid in the house. Love to work on something which impacts many in useful way. But no clue what it is.

  3. I am passionate about learning new things particularly in the space of technology. Spend a lot of time reading about distributed computing , analytics , machine learning and now deep learning. Apart from that I love cooking Indian food.

  4. Not sure.

Here’s my linkdin profile . Love to connect with you all.

1 Like

Hi everyone,

  • How did you find out about the course? I found out about part 1 of the course from HackerNews, and watched all the episodes eagerly. Before that I took the Coursera ML course with Andrew Ng, and was able to get to 3rd place in a traffic light recognition contest. The competition took model size into account (since the use case was for mobile application) and I am still a little intrigued as to what I could have done to increase the classification accuracy.
  • What I’m hoping to do with deep learning better understand the scope of problems that ML can solve, and tune my way of thinking about such problems.
    • Improve on current methods for drug discovery, like docking by Shroedinger
    • Create a monitoring app which detects audio stress signals for babies and elderly people.
    • Create a “live accompanying band” (a la Wavenet?) to play along with
  • Other tidbits: My BSc is in Biotechnology, and I am a freelance consultant for backend/frontend/mobile.

Hello together,

I am Benedikt from Germany (Düsseldorf) and I am working there in a consultancy company in the advanced analytics department.

  • How I find the course I watched Jeremy’s TED Talks and other talks online, after I read the article about MOOC Part I. After the announcement of Part II, I wanted to join it directly.
  • Hoping to do with deep learning? I want to participate in kaggle competitions to gain more practical experience + and develop new models. In addition I want to contribute to the deep learning community. I still struggeling with learning all concepts, which are really fast envolving (every week new papers are published). I think deep learning is a groundbreaking technology and has to be distributed easier to everyone.
  • It may sounds too much, but I am passionate about programming and machine learning. I like to see the results of a program/see the program working. Except of that, I like travelling a lot.
  • I worked in some Berlin startups and like the entrepreneur spirit

@jeremy & @rachel : In addition I want to thank you (like many others) for the great part I and that you continue the series with part II. I think you should consider even a part III & IV (I asked it yesterday in the livestream test).
(Sorry for the off topic):

Like @Jonas I did multiple MOOC for machine learning / deep learning (coursera, udacity, etc.). I see the big problem and need to have a continues series for a big topic like deep learning. Each MOOC starts with the basics again (hot-n encoding, dense layer, CNN, always a short intro in RNN) and just add some new/extra information. But I really miss a structured next step - like the long idea list for Part II.

I was already impressed of the Part I - many MOOC doesn’t explain the difference between CNN, ResNet, Inceptions - really a great job!!!

I didn’t want to wait for the official release and looking forward to learn with you all the concepts of PART II. (Even I have to wake up at 3:30am and take the class from 3:30am to 6am every Monday for the next 7 weeks :slight_smile: )



@burgalon, like your idea about the Wavenet band. I had some fuzzy thoughts about recognizing or creating music with DL in mind as well. Have you got prior experience with music analysis using ML / DL?

Hello everyone!

My name is David and I’m currently in the last year of my radiology residency in NYC and will be starting a fellowship in neuroradiology and imaging informatics in July.

Advances in convolutional neural networks for medical image classification are very relevant to my specialty (enough that Geoff Hinton is telling hospitals to stop training us), so have been trying to stay ahead of the curve. I’m currently competing in the Digital Mammography DREAM Challenge and plan to compete in the DSB as soon as I assemble my new home server.

I very much enjoyed part I and am looking forward to getting into the more advanced networks in part II.


Hey @iNLyze. Unfortunately I don’t have any former experience with music and DL. Re Wavenet, maybe something like MIDI (or 1980 MODs :slight_smile: ) is less CPU intensive and could achieve a more realtime solution…

Hey everybody, I am Harsh. I am an international student(Undergraduate, 2nd year, Computer Science) studying at Rochester, NY.

  • I found this course from flipboard. I used a ton of things from Part 1 to make a grocery classifier. It is still an ongoing project. Its a health app that will track what food you put in/took out of a fridge using a camera.

  • I was always fascinated about machine learning. Never really understood how it works before this course. Now I dream about it. So I plan to do anything and everything I can , with ML. I always wanted to make a clone of , an NLP company that Facebook acquired.

  • I am passionate about learning new things, especially in the field of technology. I tend to get a little cranky over optimising little things.

  • I have an eraser collection consisting of over 500 different kinds of pencil erasers.


Hi everybody !

I am very happy to be here and take this course with you as an international fellow.
I am a computer scientist, network engineer and civilian at the Ministry of environment in Ivory Coast.
Very interested in Cognitive Computing Research, I had followed several courses in Machine Learning with Udacity (still in progress with the Machine Learning Engineer Nanodegree) and recently with Kadenze where I earned the certificate of “Creative Applications of Deep Learning with Google’s Tensorflow of Parag K. Mital”. I wrote a publication on medium which details what I did for my open-ended final project.

I would like to learn the latest techniques not only for professional goals (reach the Senior Cognitive Computing Researcher) but also to well perform some personal projects including COMPETITIONS.
Indeed, I have a personal open source project for a voice or audio video translation platform (you should find more details on my github HERE). I think I could do it combining some actual methods.
But with the latest cutting-edge techniques we will learn in this course, I should be able to do it taking into account these new trends, and perhaps build it using only one neural network (that would be the ideal thing) …
So I will be happy if I could find here some friends to work on with me (It is an Open Source Project).



@Kjeanclaude I enjoyed your writeup! We’ll be looking at GANs on Monday, so you’ll be very familiar with the topic :slight_smile: Although we’ll be looking at a brand new paper

1 Like

Hi everyone !

My name is Sahil. I am working as a data scientist at a social sector startup in India “farmguide”. My work involves using deep learning to analyze satellite imagery to help the government decide which villages and crops etc. need urgent attention.

  • I’m originally from Phagwara. Its a small town in India and quite boring. But now living in the suburbs of Delhi, a city called Gurgaon which is quite a fun place.

  • I follow many deep learning researchers on twitter. I found out about course 1 from twitter itself. I knew about Jeremy because I had been following the work of Enlitic closely. And I knew Jeremy was a kaggle top ranker. So decided to do the course when it came out. My fun project from using part-1 knowledge was a neural styler. And because I totally loved the part 1, I decided to apply for deep learning part 2 as an international fellow.

  • I am passionate about deep learning because of its possibilities. Face recognition alone can make the world crime free. And it can be transformative for medicine, credit modeling etc both of which can transform the lives of poor people in developing countries. I primarily want to create deep learning models that can help improve the lives of poor people in India. I believe medicine, credit modelling, fraud detection, satellite imagery analysis can help make the lives of poor people in India and elsewhere much better.

  • I am passionate about reading mystery novels, ML research papers, listening to music, traveling to good places, music concerts and about using ML to solve impactful problems.

  • I read ML papers, kaggle kernels almost all the time. That’s something most people don’t know about me.