Chicago - USA area


#41

This Sunday 4 PM is good for me.

If not Sunday then How about Saturday for meetup?


(Sasha) #42

Doodle for scheduling to avoid a long comment chain

Ignore the dates this is just for time availability

So far, best choices are Thursday 7-9pm, Saturday/Sunday after 4pm.


(Daniel Armstrong ) #43

@agoldina, @tamhash @larcat

if you can make it here is information on our Hangout-Meet Today @ 4pm

Join by phone

‪+1 413-300-1551‬ PIN: ‪555 913 710‬#


#44

Thanks @Daniel.R.Armstrong !


(Sasha) #45

First google hangout went well! I enjoyed talking with you @tamhash and @Daniel.R.Armstrong

Next hangout will be March 3rd at 4pm. The goal is for us all to put in some work on one project and spend ~10 presenting at the next hangout. I’ll be reviewing the classes and applying classwork to different datasets as well as starting the Toxic Comment Classification challenge on Kaggle.

Things I have mentioned in today’s discussion to share:

Podcast about the Zillow Zestimate

Label my Text Widget


#46

hi @agoldina @Daniel.R.Armstrong,

Great to talk with you.

Here is study group I mentioned during our discussion

This Week in ML & AI Meetup
Workspace URL: twimlai.slack.com

They have been through two lectures yet and uploaded video for first session of meet-up on youtube.


(Daniel Armstrong ) #47

@tamhash and @agoldina It was great to meet you both, and learn about the projects you are focusing on.

@agoldina I would love to work with you on aspect based sentiment analysis of review data after you finish the toxic comments challenge. What do you think about using review data from amazon/yelp then use that pre-trained model model on yelp moving company data.

@tamhash I would love to work on mri data with you. I found some data-sets/resources that might help

NLP/Aspect based sentiment analysis:
Text summarization, lda2vec, aspect based sentiment analysis(blog), aspect-analysis(monkeylearn.com), aspect-based-opinion-mining(medium),topic modeling(nlpforhackers), Amazon review summarizer(tensorflow), aspect-based-sentiment-analysis-competition(dataset), state and local financial data(tax policycenter.org)
senti aspect(notebook)

Lexicons:
I lost where I put all my lexicons, if/when I find them I will let you know, my favorite on is from Saif Mohammad(NRC-Emotion-Lexicon). Here is a few more.

I am thinking it might be a good idea to use text simplification for data reduction and a way to adjust for authorship bias(collaborative filtering).

Sarcasm can cause all sorts of issues so a corpus might help:

Government Data:
I found this when I was wanting to find out why Illinois was so broke, but I never really used it.

Medical imaging:


#48

@Daniel.R.Armstrong Thanks!


(Sasha) #49

Haha why Illinois is so broke. That could be a fun one. Maybe because they sell off all good sources of income such as parking and politicians are very corrupt?


#50

Hi @Daniel.R.Armstrong and @agoldina,

I found this very good article about medical imaging with fastai

He also gave some links for publicly available medical imaging datasets.

He also developed a library on the top of fastai.

I found one more similar and very good repository for medical imaging using fastai


(Daniel Armstrong ) #51

I read the first article you posed, and I thinking my knowledge about medical imaging might not be good enough to munge the data into a format we can use, did you have any success? What data set/type of images sould/did I/we/you start with. I am very curious about the alzheimer data set(kaggle).

Do we still want to have a conference tomorrow? I would love to learn more about medical imaging.

I have spend most of my time on a ds certification and teaching/preparing for the meetup I lectured at last Tuesday. I also learned about Broadcasting.


#52

We can have conference. You can explain about broadcasting if you want. If you have some other topics to discuss then you are most welcome to do so.

I spent last month with histopathology kaggle competition without much success. But I learned about “gradient heat map” while working on this competition. I can talk about that.

I haven’t done with medical imaging yet but planning to take on one project this month. I can talk about my plan too.


(Daniel Armstrong ) #53

Sounds good to me, do you want to meet at 4pm? @agoldina


(Pierre Guillou) #54

Hello, I just started a discussion thread on fastai study groups to gather feedbacks from the organizers and participants in order to list the best practices and also to avoid some gaps.

You will find more information in this post. Thank you if you can take a few minutes to participate in the discussion :slight_smile:


#55

4 pm is good for me. I hope @agoldina can join us too.


(Daniel Armstrong ) #56

I am have a car issues, I wont be able to meet up today, sorry


#57

Sure. How about your availability in coming Sundays ?


(Daniel Armstrong ) #58

I cant do the next three Sunday. :frowning:


#59

Ok. We will see if next month is possible.


#60

anyone got invitation for international fellowship (online) 2019 part 2 ? If yes then interested in reinvigorating Chicago study group?