Rishabh Iyer
@rishiyer
Followers
2K
Following
2K
Media
36
Statuses
560
Prof. at UTD CS, Director @caraml_lab | ML/AI/Optimization | Ex-Microsoft | MS, PhD: UW, BTech: IITB | https://t.co/TB3cpJcBeC
Dallas, TX
Joined February 2016
I just finished up a new course I've been teaching for Spring 2021 titled "Optimization in Machine Learning". Different from typical "OptML" courses, I covered both discrete and continuous optimization in 11 weeks. Here is the youtube playlist: https://t.co/ap0JYMGfrQ.
youtube.com
7
216
977
Takeaway: Your split strategy defines the story your evaluation tells. Random splits answer “can the model generalize on average?” Temporal splits ask “can it predict the future?” Stratified splits ensure fairness across classes. Group-based and leave-one-group-out splits measure
0
0
0
To ground these different split strategies in practice, we ran four small experiments using synthetic datasets. Each experiment highlights how the wrong split can give misleading confidence in your model — and how the right split exposes the truth. The Figure below shows the
1
0
0
ML Nugget #2: Choosing the Right Train/Test Split Splitting your dataset into training, validation, and test sets feels like one of the most straightforward steps in machine learning. But the truth is: how you split the data can dramatically change your evaluation and, more
1
0
2
To make this concrete, I ran two simple simulations that show just how dangerous distribution shift can be. Feature Shift (Covariate Shift): Here, the underlying input features gradually drift over time. Think of stock price features that evolve as market conditions change, or
1
0
0
ML Nugget 1: Beyond Train/Test: The Deployment Gap and How to Quantify It Every ML 101 course teaches you about train, validation, and test splits. The train set helps the model learn, the validation set helps tune hyperparameters, and the test set estimates generalization. We
1
0
6
I’ve been teaching AI/ML courses for several years and working with companies and startups for over a decade. Along the way, I’ve picked up practical lessons in machine learning that often don’t make it into standard textbooks or courses. I’m starting a short series to share
0
0
8
Please congragulate my dear friend @Sriraam_UTD who is now a AAAI fellow. Well done @Sriraam_UTD . Details here:
aaai.org
Listing of elected AAAI Fellows. Their accomplishments range from advances in the theory of AI, to unusual accomplishments in AI technology and applications.
8
3
26
Very nice insights! Agree 100%!
7 lessons from AirSim: I ran the autonomous systems and robotics research effort at Microsoft for nearly a decade and here are our biggest learnings. A thread 🧵 Complete blog:
0
0
0
Congratulations India! What a World Cup win! Indian team today is one of the strongest it has ever been! While India has had good batsmen always, India’s bowling has really improved. This has been a game changer!
0
0
14
Today I'm filled with joy to see the Ram Mandir opening in Ayodhya!! Lord Ram exemplifies what it means to be an ideal person - ideal king, ideal son, ideal husband, and ideal in every way! I have little doubt that the next decade will be that of India! #RamMandirPranPrathistha
0
0
15
Thank you for the invitation! I presented work done by @krishnatejakk's Ph.D. on subset selection for compute-efficient deep learning! I also enjoyed all the other talks at the conference! It was a solid program! Congrats to the organizers of @indoml_sym!!
Day 1: Session 2 Machine Learning Talk 2: Rishabh Iyer Professor UT Dallas Subset Selection for Compute-Efficient Deep Learning Professor Rishabh took us through approaches like GLISTER, GRAD-MATCH, MILD, which helped us make our concepts on subset selection crystal clear!
0
1
14
Day 1: Session 2 Machine Learning Talk 2: Rishabh Iyer Professor UT Dallas Subset Selection for Compute-Efficient Deep Learning Professor Rishabh took us through approaches like GLISTER, GRAD-MATCH, MILD, which helped us make our concepts on subset selection crystal clear!
0
5
14
Excited to be presenting a talk at IIT Bombay for IndoML 2023! IndoML 2023 will be from 21 - 23 December 2023!
IndoML 2023 brings a series of exciting talks on state-of-the-art ML Technologies. Dr. Rishabh Iyyer of UT Dallas will discuss “Subset Selection for Compute-Efficient Deep Learning: Orders of Magnitude Speedups with Subsets of the Training Data!” #indoML #mlsymposium #indoML2023
0
2
12
Exciting work in collaboration with Adobe Research! We used subset selection to make training of Large Language Models faster! https://t.co/tAzxR91omE
#EMNLP2023 Paper announcement (Findings). We show the first application of data subset selection techniques for efficiently training language models. Wonderful collaboration with Kowndinya Renduchintala, @krishnatejakk, @rishiyer , @ganramkr Milan Aggarwal, @KbalajiTweets
0
2
36
Our work, in collaboration with Prof. Abir De of IIT Bombay, @eeshaan_jain, @rishiyer and @fooobar, on Learning to Select a Subset of Training Examples to Generalize Efficient Model Training has been accepted at #NeurIPS 2023. (1/n)
3
7
136
In short, the story of economic management, between the Congress and BJP…
409
3K
7K
India's successful lunar mission (landing its Chandrayaan-3 spacecraft on the moon) is another one of many straws in the wind showing its ascendence. As previously shown in my health index for countries, which is used to derive my projections for countries' next 10-year growth
195
2K
10K
Very nice work by @tiwarishabh16 and my close friend and collaborator @doktorshenoy on removing irrelevant features for mitigating bias!
🚨 Excited to share our #ICML2023 work on the Feature Sieve, by which we automatically identify and suppress irrelevant or spurious features in deep networks, hence mitigating simplicity bias. Paper: https://t.co/0od1W0soAs
0
0
4
We can consider the mathematical beauty of problems, come up with theoretically principled approaches and not so cool research areas where there is still scope of making impact! Not everything needs to be solved with sheer scale!
0
1
9