Lester Mackey

American computer scientist and statistician From Wikipedia, the free encyclopedia

Lester Mackey

Lester Mackey is an American computer scientist and statistician. He is a principal researcher at Microsoft Research and an adjunct professor at Stanford University. Mackey develops machine learning methods, models, and theory for large-scale learning tasks driven by applications from climate forecasting, healthcare, and the social good. He was named a 2023 MacArthur Fellow.[1]

Early life and education

Mackey grew up on Long Island.[2] He has said that, as a teenager, the Ross Mathematics Program in number theory introduced him to proof-based mathematics, where he learned about induction and rigorous proof.[2] He got his first taste of academic research at the Research Science Institute.[2] He joined Princeton University as an undergraduate student, where he earned his BSE in Computer Science. There he conducted research with Maria Klawe and David Walker.[3] Mackey was a graduate student at the University of California, Berkeley, where he earned a PhD in Computer Science (2012) and an MA in Statistics (2011).[1][4] At Berkeley, his dissertation, advised by Michael I. Jordan, included work on sparse principal components analysis (PCA) for gene expression modeling, low-rank matrix completion for recommender systems, robust matrix factorization for video surveillance, and concentration inequalities for matrices.[5] After Berkeley, he joined Stanford University, first as a postdoctoral fellow working with Emmanuel Candès and then as an assistant professor of statistics and, by courtesy, computer science. At Stanford, he created the Statistics for Social Good working group.[1]

Research and career

In 2016, Mackey joined Microsoft Research as a researcher and was appointed as an adjunct professor at Stanford University. He was made a principal researcher in 2019.[1]

Mackey's early work developed a method to predict progression rates of people with ALS. He used the PRO-ACT database of clinical trial data and Bayesian inference to predict disease prognosis.[1] He has also developed machine learning models for subseasonal climate and weather forecasting, to more accurately predict temperature and precipitation 2-6 weeks in advance.[1] His models outperform the operational, physics-based dynamical models used by the United States Bureau of Reclamation.[1]

Awards and honors

Selected publications

  • Luke de Oliveira; Michael Kagan; Lester Mackey; Benjamin Nachman; Ariel Schwartzman (July 2016). "Jet-images — deep learning edition". Journal of High Energy Physics. 2016 (7). arXiv:1511.05190. Bibcode:2016JHEP...07..069D. doi:10.1007/JHEP07(2016)069. ISSN 1126-6708. OSTI 1271300. S2CID 30627853. Wikidata Q123016814.
  • Neil Zhenqiang Gong; Ameet Talwalkar; Lester Mackey; Ling Huang; Eui Chul Richard Shin; Emil Stefanov; Elaine (Runting) Shi; Dawn Song (April 2014). "Joint Link Prediction and Attribute Inference Using a Social-Attribute Network". ACM transactions on intelligent systems and technology. 5 (2): 1–20. doi:10.1145/2594455. ISSN 2157-6904. S2CID 7277785. Wikidata Q123016825.
  • Lester W. Mackey (2009). "Deflation Methods for Sparse PCA" (PDF). Advances in Neural Information Processing Systems 21. Advances in Neural Information Processing Systems. Wikidata Q77680580.

References

Loading related searches...

Wikiwand - on

Seamless Wikipedia browsing. On steroids.