Assistant Prof @umasscs, @UMassAmherst. Formerly postdoc @Stanford CS. PhD from @mldcmu. Research: ensembles, structured sparsity, deep fusion, ML4Health.
Amherst, MA (socially distant)
Joined November 2015
- Tweets2.5K
- Following286
- Followers976
- Likes1.3K
Pinned Tweet
Thank you for funding 4Thought, our project on early forecasting of Alzheimer’s disease from brain MRIs + multimodal patient data!
.@UMassIALS has named six campus research teams as recipients of the first Manning/IALS Seed Grants that will support new product research and services to benefit human health and wellbeing. umass.edu/newsoffice/article…
Madalina Fiterau retweeted
I want to clarify that this does not apply to the UMass Amherst CS department. Our PhD admissions and faculty hiring processes are continuing as usual.
Very sad to hear that UMass is rescinding all PhD admissions for Fall 2025. This means there will be no new PhD cohort, and UMass will not have a graduating class that year. This is just unprecedented and I am so very sorry for all the affected students.
Readers added context they thought people might want to know
This is incorrect. UMass is NOT rescinding all PhD admissions for Fall 2025. Only the Morningside Graduate School has done this. Other UMass programs have not.
Read here for details: masslive.com/worcester/2025…
“applicants to UMass other graduate schools …. are not impacted”
boston.com/news/health/20…
Madalina Fiterau retweeted
The Machine Learning for Healthcare (MLHC) 2025 Call for Papers is out!
mlforhc.org/paper-submission
Pre-submission intent deadline: April 4th, 2025
Full submission deadline (Research and Clinical Abstracts Tracks): April 11th, 2025
Submit via OpenReview: openreview.net/group?id=mlfo…
Madalina Fiterau retweeted
🎉 #PhDone
After years of research, debugging marathons & existential crises, I defended my PhD on Adaptive Deep Learning for Personalized Time-Series Modeling!
Huge thanks to my advisor @MFiterau , my committee, & amazing collaborators!
On to the next adventure! 🚀
#PhinisheD
Our lab's contributions at @NeurIPSConf workshops include treatment disparity detection, LLM-augmented experimentation and RAG-assisted zero-shot forecasting! The awesome lead authors of the papers Iman Deznabi (@IDeznabi), Purity Mugambi and Shiv Shankar are on the job market!!
Now at the #ML4H2024 Symposium poster session: Shiv Shankar is presenting his research on estimating distributional and risk-aware measures of treatment effects under interference. Joint work with Ritwik Sinha. Project sponsored by @AdobeResearch
Also at the #ML4H2024 Symposium poster session: Purity Mugambi presenting her LLM-based method for Automated Cohort Extraction from Large EHR Databases. (Useful in performing the multi-dataset analysis showed in the WiML poster).
Madalina Fiterau retweeted
Join us today for the WiML Poster Session, co-hosted with the {Dis}Ability and Queer in AI Workshop and the LatinX in AI (LXAI) Research Workshop in the Soltz foyer! Any ICML registration is acceptable for entry. Register for the #WiML Symposium here: buff.ly/3WRW8oZ
Madalina Fiterau retweeted
We have open-sourced the code and data for "Improved Zero-Shot Classification by Adapting VLMs with Text Descriptions" @CVPR at: github.com/cvl-umass/AdaptCL…
Also, find the poster and video links at the GitHub repo!
Humble to announce my *ONE* and only paper arxiv.org/abs/2401.02460 got accepted :))) @Grant_Van_Horn @MajiSubhransu
Poster: A/B Testing under Interference with Partial Network Information
Are your A/B tests affected by interference? Use our new variance-reducing estimator UNITE which works under a non-linear motif model! Learn more at Poster 65, Friday May 3rd @aistats_conf #AISTATS24 proceedings.mlr.press/v238/s…
Are your A/B tests affected by interference? Use our new variance-reducing estimator UNITE which works under a non-linear motif model! Learn more at Poster 65, Friday May 3rd @aistats_conf #AISTATS24 proceedings.mlr.press/v238/s…
Madalina Fiterau retweeted
Fun work done in collaboration with my PhD Gustavo Perez and Dan Sheldon on obtaining reliable count estimates from computer vision for various tasks. This won the best paper for the social impact track at AAAI 24.
People count things accurately, but AI counts faster. By having #AI and humans work together, researchers @manningcics have made counting buildings in disaster zones more efficient than ever. 👨 🤖 #ArtificialIntelligence #UMass bit.ly/4aLybmI
Madalina Fiterau retweeted
Here is a list of contributions from the @manningcics Center for Intelligent Information Retrieval (CIIR) to @SIGIRConf 2024!
Multiple exciting papers on RAG, Generative Retrieval, Proactive Conversations, and LLM Personalization.
Pre-prints coming up, stay tuned... #SIGIR2024
Madalina Fiterau retweeted
💻The code for Branched CALM-Net is available here: github.com/Information-Fusio… [8/8]
#StressPrediction #DeepLearning #AdaptiveDeepLearning #TimeSeries #Research
Madalina Fiterau retweeted
🧠 Introducing Branched CALM-Net: a groundbreaking deep learning framework for predicting stress from mobile sensor data. Our approach uniquely combines dynamic clustering with branched deep learning, offering unprecedented personalization. [2/8]
Madalina Fiterau retweeted
📈 With just 1 week of data, Branched CALM-Net surpasses the state-of-the-art by a 17.24% improvement in F1-score, marking a significant advancement in personalized stress prediction technology. [3/8]
Madalina Fiterau retweeted
🎉Thrilled to share our latest paper "Dynamic clustering via branched deep learning enhances personalization of stress prediction from mobile sensor data" published in Nature Scientific Reports! Co-first-authored with @yunfei_luo . 🎉 [1/8]
nature.com/articles/s41598-0…
Madalina Fiterau retweeted
MLHC 2024 Call for Papers is now open!
mlforhc.org/paper-submission