How has your journey been so far, learners?


(Kevin Bird) #144

Definitely @jeremyphoward and @math_rachel. I also really like @Smerity.


(Vishal Pandey) #145

I too got a Handle…! Vishutcyrus


(Jeremy Howard) #146

If you check out my favorited tweets (I’m jeremyphoward) you can see who tweeted the stuff that you find interesting there, and follow them. Also, be sure to follow me, so I can DM you on twitter. Welcome to the twitter ML community!

And @kevinb has excellent taste…


(helena s) #147

twitter is my main source of all news ML/DL/CV - also the best for political quips…
oh, and i am @glagolista there


(ABILASH AMARTHALURI) #148

Hello everyone… I am a Software Engineer by profession looking to step into the Deep Learning world. A big thanks to @jeremy and @rachel for this course. I am doing my exploration on Data Science by surfing the web here and there, attended a workshop @ Stanford etc., But nothing beats this course. I am deeply inspired by how @jeremy shows the best techniques to achieve state of the art results. As a software engineer I have worked on some interesting problems like Query understanding and Title optimization using computer science techniques like Fuzzy String matching, Language models etc., but felt like i can tackle these problems much better with knowlege of Deep Learning.

Apart from professional work, I play Drums in a small church band, but currently we do not have a Electric Guitar player in our band. I hope Deep Learning can be my Electrician Guitar player if trained on many songs using RNNs :thinking:

I am so thankful for this amazing opportunity and looking forward to learn a lot.

—Abilash


(rachana) #149

I am a software engineer. Primarily writing code using Java. For last few year I transitioned into the big data space. I am doing machine learning using distributed environment using big data toolsets. After learning about Machine Learning I want and knowing the pain of creating feature engineering.

I wanted to explore the shortcut of feature extraction by deep learning. I am new to Python, and some of the Deep Learning concepts are overwhelming for me at this point…

Want to design real time solution using Deep Learning in the distributed setting.


(Sanyam Bhutani) #150

Hello World,

I’ve finally found the courage to introduce myself after the beginner’s category was created and I felt that I wasn’t the only one in uncharted waters.

I’m a CS Undergrad student from India.
I’d like to share how I want my journey to be from here (what I hope I’ll be able to target after completing part 1,2) :slight_smile:

What I achieve to get out of deep learning is, I wish to help improve Agriculture in my country. Here’s some background: Labour, Agriculture are the roots of my Nation’s economy. I wish to help improve agritech by helping analysing and suggesting better methods/crops and all the possibilities.

I’m also fascinated by Computer Vision. I want to help replace ID cards by Face IDs at my college, create a automated theft detector for retails stores.

Also, My country consists of many small scale retailers. I’d like to help them by analysing sales and other details.

I want to confess that I have almost no idea as to how I’m gonna go about doing this, but I have faith in our FASTAI community that you guys will definitely help me on my path to these goals. Also now that I’ve mentioned these. There is no going back :blush:

Regards,
Sanyam Bhutani


(Maureen Metzger) #151

@init_27 (Sanyam) – don’t worry about how you will get there, just take the next step as it comes!

And I would say that to all here who aren’t sure how they are going to get to where their dreams are leading them – kudos for taking the first step, and just keep on moving forward, step by step. And try to enjoy the process :slight_smile:

I say that with the wisdom of my 53 years :joy: Life will surprise you, sometimes not pleasantly, but all in all it’s an amazing journey!


(Tuatini GODARD) #152

@init_27 very good projects you have! I see a lot of people wants to do something with deep learning/agriculture. I’m also interested by these domains :slight_smile: . This project really got me excited as you can retrieve all the data from a plant and maybe predict the best parameters for a given plant to grow it? That would be amazing to see people from this community making a startups/projects which are really impactful on domains that matters like bioinformatics/agriculture/health etc etc…

And I would say that to all here who aren’t sure how they are going to get to where their dreams are leading them – kudos for taking the first step, and just keep on moving forward, step by step. And try to enjoy the process :slight_smile:

Thanks a lot for these kind words @memetzgz . I left my job 1 year ago to switch to Data science and last week I’ve taken my first freelance mission in ML that I totally failed :laughing: . But I was just not ready and as you say life is full of surprises. Data science is just a domain where you shouldn’t be too much worried about opportunities showing up. If you keep working hard (and smart) I’m sure opportunities will come to you at some point :slight_smile:


(Maureen Metzger) #153

@ekami I know what you mean about just not ready. I botched a data scientist interview a couple of months ago by failing an entry-level coding problem. Oops :grimacing: But life goes on. And as you said, more opportunities will come our way. Patience is a useful muscle to exercise!


(Sanyam Bhutani) #154

Thank you for you kind words!
I’ll keep trying :slight_smile:


(Sanyam Bhutani) #155

It’s amazing. I think it is the end goal for all of us practitioners, to create something that might just be able to run even on a Raspberry Pi board. The most underpowered Motherboard that one can get hands on, but one that’d be accessible to all.

I’ll try my best :slight_smile:


(Michael Moret) #156

Hey everyone,

I’m a bioengineering student from Lausanne, in Switzerland. I’m currently working on my master thesis in Boston and I’ve been fascinating by deep learning since a couple of years.

I tried to use deep learning on different problems until now, such as creating a poker agent or predict cancer severity from images. For this last project, the first version of the fast.ai course help me a lot. Really a lot. Particularly, I really enjoyed the way @jeremy presents his course – even if I’m a full-time student, I usually never go to class because I hate the way they are given. I think someone living in a galaxy far, far away said a long time ago «If you don’t use a Top-Down approach to teach, you will become an agent of evil».

Now, I’m working notably on protein engineering with DL and image analysis of cell culture. My goal is to get into a PhD program in ML next year.

If anybody is living near Boston or Lausanne, please drop me a message!

–Michael (Twitter: @moret1788)


(Jeremy Howard) #157

We’ve had one student have to drop out unfortunately. So we’ve added a new participant, @lrajlich (Luke Rajlich). Luke was a co-founder of the very popular Farmville game (I’m sure some of you played it! :slight_smile: ) and tells me he’s been busily catching up on the lesson videos over the weekend… Welcome Luke!


(Hiromi Suenaga) #158

I’m also curious to know what you’ll be working on :slight_smile:


(Kevin Bird) #159

Welcome @lrajlich!


(Asir Saeed) #160

Thinking about doing something with word vectors on the Aozora Bunko corpus. Nothing too specific yet. :slight_smile: Any input is appreciated! :grin:


(Anand Saha) #161

@lrajlich How about Deepville this time? :wink: Welcome!


#162

@lrajlich I would be interested in your thought on how deep learning will help startup founder as most of the startup don’t have the data to start with. Would love to know your perspectives.


(Luke Rajlich) #163

Thank for the intro @jeremy and thanks @zaoyang for the intro here. Happy to be a late comer, I want to get my hands dirty with deep learning and building models. I don’t have an extremely specific objective other than to learn how to practically apply deep learning and I’ll go from there once I have more experience.

I’ve had a long passion for analytical work and some experience experience - first job out of college was working on quant trading strategies at a hedge fund and I built a rather janky analytics system to measure cohort retention in R – this was pre mixpanel or other SaaS analytics providers at my startup with Zao.

Thanks @KevinB! @anandsaha you jest, but maybe ;)… @atulkum - it probably depends on what the startup is doing but if you plan to build features to run models against user input data, then you could possibly train a model on publicly available data? Also having some friendlies who fit your customer profile and are willing to share data with you would be a huge help to get the ball rolling.