Jagdeep Bhatia — سيرة طفل معجزة

السيرة الذاتية

Second place, and $175,000

Society for Science lists Bhatia second in the 2020 Science Talent Search, behind Lillian Petersen and ahead of Brendan Crotty, with an award of $175,000. His school is given as Watchung Hills Regional High School in Warren, New Jersey, his home town as Green Brook in the same state.

The competition that year drew 1,993 entrants, from whom forty finalists were chosen; every finalist received at least $25,000, and the top ten took between $40,000 and $250,000. Second place is therefore not a consolation. It is the second-largest award in what is generally treated as the most demanding pre-university research competition in the United States.

The idea: let the learner choose the question

Bhatia's project is titled Simple and Fast Algorithms for Interactive Machine Learning with Random Counter-Examples. The Society describes it as developing fast machine learning algorithms for computer programs that learn new concepts from an instructor, asking strategically selected questions rather than random ones, with the aim of training robots and automated devices more efficiently.

The distinction matters more than it sounds. In ordinary supervised learning, a model is handed a pile of labelled examples and has no say in which ones it sees. In the interactive setting, the learner can put a question to a teacher, and the choice of question determines how fast it converges. A well-chosen counter-example, one that contradicts the learner's current guess, can be worth thousands of random samples. Algorithms that are both simple and fast at generating those questions are the practical bottleneck between the theory and a robot that can be taught on a factory floor in an afternoon.

Where it sat in the 2020 field

The 2020 top ten was unusually wide in subject. Lillian Petersen took first with $250,000. Brendan Crotty, a home-schooled student from Muskogee, Oklahoma, took third and $150,000 for a hybrid industrial gas burner that emitted 19 per cent less nitrogen dioxide. Rupert Li was fourth, Anaiah Thomas fifth, Katherine St George sixth, Alek Westover seventh, Adriane Thompson eighth, Rohan Wagh ninth and Arjun Neervannan tenth.

Machine learning entries have become a fixture of the search, and the reason is the same one that makes them contentious: the equipment is a laptop. A student with mathematical ability and a public dataset can produce work at the edge of a research literature without a bench, a supervisor or a grant. That widens who can compete, and it raises the bar for what counts as original.

What the prize is for

The Science Talent Search does not judge a demonstration. Entrants submit a full research report along with their school record and essays, and the forty finalists are then interviewed in Washington by panels of working scientists who probe the surrounding field as much as the project itself. A finalist who cannot explain why the problem matters, or what would falsify the result, does not place.

Against that standard, an award of $175,000 is a statement about the student rather than the paper. The Society's own framing treats the search as identifying future research scientists, and the roll of past finalists is heavy with people who became exactly that.

الاختبار المجاني

وكم يبلغ ذكاؤك؟

عشرون سؤالاً وثماني دقائق على الساعة، ونسبة مئوية مقاسة مقابل كل من خاضه. دون تسجيل.

اخوض اختبار الذكاء →

What is documented, and what is not

Firm: Bhatia placed second in the 2020 Regeneron Science Talent Search and received $175,000; he attended Watchung Hills Regional High School in Warren, New Jersey, and is listed from Green Brook, New Jersey; his project concerned simple and fast algorithms for interactive machine learning with random counter-examples.

Not documented, in the source used here: his age or birth year, where he went to university, whether the work was published in a journal or conference, and what he has done since March 2020. The Society's finalist listing is precise about the project and silent about the person, which is the reverse of most prize coverage and, for a page like this one, the better half to have.

الجدول الزمني للإنجازات

Jagdeep Bhatia — every milestone on this page, in order.

2020

تم اختياره كأحد الأربعين متأهلاً للتصفيات النهائيةتم اختياره كأحد الأربعين متأهلاً للتصفيات النهائية في مسابقة ريجينيرون للمواهب العلمية، من بين 1,993 مشاركاً

Mar 2020

احتل المركز الثاني وحصل على 175,000 دولاراحتل المركز الثاني وحصل على 175,000 دولار، خلف ليليان بيترسن

2020

نشرت جمعية العلوم مشروعهنشرت جمعية العلوم مشروعه حول التعلم الآلي التفاعلي مع حالات مضادة عشوائية

Since 2020

لم يتم العثور على سجل منشور إضافيلم يتم العثور على أي سجل منشور إضافي في المصادر المستخدمة لهذه الصفحة

Watch: Jagdeep Bhatia on camera

Zero Robotics - Jagdeep BhatiaStorming Robots

مقارنة تاريخية

الشخص	الدولة	الإنجاز	العمر / البيان
Jagdeep Bhatia	الولايات المتحدة	المركز الثاني، مسابقة ريجينيرون للمواهب العلمية 2020	—
Melissa Rey	الولايات المتحدة	الجائزة الكبرى، تحدي ديسكفري إديوكيشن 3M للعلماء الشباب 2008	—
Sara Volz	الولايات المتحدة	المركز الأول، مسابقة إنتل للمواهب العلمية 2013	—
Jonah Kallenbach	الولايات المتحدة	المركز الثاني، مسابقة إنتل للمواهب العلمية 2013	—
Victor Cai	الولايات المتحدة	المركز الثاني، مسابقة ريجينيرون للمواهب العلمية 2022	—

What is documented — and what is not

✓ Documented

Second place and $175,000 at the 2020 Regeneron Science Talent Search
Attended Watchung Hills Regional High School, Warren, New Jersey; listed from Green Brook
Project title and description published by Society for Science
Forty finalists were chosen from 1,993 entrants in 2020

? Unverified, disputed or wrong

Age and birth year are not stated
No record here of the work being published in a journal or conference
Nothing documented about his university or career after March 2020

Questions people ask about Jagdeep Bhatia

What did Jagdeep Bhatia win?

Second place and $175,000 at the 2020 Regeneron Science Talent Search, according to Society for Science.

What was his project about?

Fast algorithms for interactive machine learning: letting a program that is learning a new concept ask strategically chosen questions instead of sampling at random, to train robots and automated devices more efficiently.