Karimi et al. win ECML PKDD Test-of-Time Award
Understand this story
Amazon senior applied scientist Hamed Karimi, along with Julie Nutini and Mark Schmidt, received the 2026 ECML PKDD Test-of-Time Award for their 2016 paper "Linear convergence of gradient and proximal-gradient methods under the Polyak–Łojasiewicz condition." According to Amazon Science, the paper proved that the Polyak–Łojasiewicz condition, long overlooked by the machine learning community, can guarantee that gradient descent converges at a linear rate, and clarified the relationships among various optimization methods under weak conditions; the paper has been cited more than 2,000 times and has entered graduate textbooks.
AI overview based on source reports · Updated 15 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Amazon ScienceHamed Karimi wins ECML PKDD Test-of-Time Award
Amazon senior applied scientist Hamed Karimi and coauthors Julie Nutini and Mark Schmidt won the 2026 ECML PKDD Test-of-Time Award for their 2016 paper on linear convergence of gradient and proximal-gradient methods under the Polyak–Łojasiewicz condition. The paper showed that the largely overlooked 1963 inequality can prove fast linear convergence for standard training methods even when many solutions work equally well, mapped how previously scattered weaker-condition results relate, and extended the analysis to nonsmooth landscapes and algorithms including coordinate, sign gradient, and stochastic gradient descent. It has been cited over 2,000 times and is used in graduate textbooks.
Activity for this story
- Current comparable activity
- 7
- Peak comparable activity
- 10Nov 6 00:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.