Sarthak
@TheDataHandler
Musing about deep learning, sports, data-viz and culture
You might like
I've often said Twitter is the place you come to see your idols fall.
Today is fast.ai’s biggest day in our four year history. We are releasing our new course, fastai v2, fastcore, fastscript, fastgpu, and the official launch of the book. 1/ fast.ai/2020/08/21/fas…
Wearing masks will help save lives & slow down this pandemic. Shortage of masks in your area? Concerned about taking supplies away from frontline workers? Don’t worry, you can make your own mask at home in minutes! Learn more at Masks4all.co #Masks4All #coronavirus
The beautiful game demands such beautiful visualizations!
We got something new! A #ConvexHull chart to measure the compactness of a #football team. All #dataprep again done in @ExasolAG. For the @tableau viz I tried to orientate on the clean style of @TriMyData. But I am sure she can still improve something. #sportsanalytics #dataviz
For a #datascience problem involving prediction of housing prices for a #USA town, there were ~80 base features to model from. A correlation matrix helped me uncover hidden structure and relationships in the #data before deploying a #predictive statistical model. #ml #analytics
With Pakistan, Sri Lanka, and Afghanistan being beaten comprehensively, subcontinent teams have had a terrible start to the Cricket World Cup. After being put to bat against a fiery South African bowling attack, can the Tigers from Bangladesh turn the tide? #CWC19 #SAvsBAN
Attending the @AIEverything1 Summit earlier this month in #dubai was a pleasure. Met @kaggle grandmasters and #ai mentors, thoroughly enjoyed a talk on GANs, solved a Simpsons image classification problem with Deep Learning, and plenty more! #dl #ml #data #future
Intriguing research on why rainbow colormaps are ineffective and mislead by discretizing continuous data. 🌈 It's attractive but comes at the cost of misrepresenting data in unpredictable ways!
We got something new! A #ConvexHull chart to measure the compactness of a #football team. All #dataprep again done in @ExasolAG. For the @tableau viz I tried to orientate on the clean style of @TriMyData. But I am sure she can still improve something. #sportsanalytics #dataviz
True, important to be mindful of what problems your model can help solve. Pursuing intellectual curiosity, however, is quite meaningful.
There is an ongoing misconception that AI/ML are intrinsically valuable, and that therefore working in the field is bound to make you rich. A ML model is only as valuable as the problem it solves. ML without an application isn't worth anything (beyond intellectual curiosity).
While solving complex #datascience problems, feature engineering can significantly improve results and set your #MachineLearning model apart. Exploratory #data #analysis or EDA can lay the groundwork for the same.
Visualizing the most expensive #soccer #football transfers in the 2018-19 preseason The transfers are broken by player position and destination league. Looks like the Seria A had a ton of transfer activity, perhaps due to the record breaking transfer of a certain #Ronaldo
A streamgraph is a stacked area graph, positioned around a central axis, rendering a flowing shape to the visualization. I am using it here to showcase counts for the top 7 medal winning regions at the @Olympics #dataviz #DataScience #DataAnalytics
Has any country dominated an Olympic sport like the #USA has dominated Basketball?! #Basketball was introduced in the Summer #olympicgames in 1936 and barring the 1980 Moscow games which they boycotted, both their men's and women's contingent have won a medal every time.
Thanks @datavisclub for organizing a live discussion with @infowetrust. As a young student of the data visualization field, I am pretty excited to engage with someone who's works have inspired me and forced me to rethink everything I knew about the subject.
The 2016 #olympicgames had the highest percentage of women participation! However, this 120 year journey to achieve #genderequality hasn't been easy and we have miles to go! As some of my #Datavisualization show, here's hoping we continue to make giant strides #feminism #women
United States Trends
- 1. #Worlds2025 39.8K posts
- 2. #TalusLabs N/A
- 3. Doran 16.6K posts
- 4. #T1WIN 26.4K posts
- 5. Sam Houston 1,564 posts
- 6. Boots 29.5K posts
- 7. Oregon State 4,850 posts
- 8. Lubin 5,740 posts
- 9. Faker 31.8K posts
- 10. #GoAvsGo 1,584 posts
- 11. Louisville 14.5K posts
- 12. Keria 10.1K posts
- 13. #T1fighting 3,385 posts
- 14. Miller Moss 1,242 posts
- 15. Emmett Johnson 2,560 posts
- 16. UCLA 7,864 posts
- 17. Oilers 5,267 posts
- 18. Hyan 1,372 posts
- 19. Nuss 5,731 posts
- 20. Frankenstein 126K posts
Something went wrong.
Something went wrong.