Candidate Information Title Machine Learning Data Science Target Location US-CT-New Haven Email Available with paid plan Phone Available with paid plan

20,000+ Fresh Resumes Monthly     View Phone Numbers     Receive Resume E-mail Alerts     Post Jobs Free     Link your Free Jobs Page     ... and much more Register on Jobvertise Free

Search 2 million Resumes Keywords: City or Zip:

Related Resumes Machine Learning Data Science Hartford, CT Machine Learning Data Analysis, programming Rocky Hill, CT Data Analysis, Machine Learning, Operations Research Storrs Mansfield, CT Solutions Architect Machine Learning West Babylon, NY Machine Learning Customer Experience Norwalk, CT Computer Science Machine Learning West Haven, CT AI - Machine Learning - Computer Vision - Python - R&D Stamford, CT

KonstantinosNikolakakis(PHONE NUMBER AVAILABLE EMAIL AVAILABLE knikolakakis.org KonstantinosNikolakakis K-Nikolakakis Research Portfolio & ExpertiseStatistical Machine Learning, Generalization Error Analysis, Risk-Aware Decision Making, Multi-Armed Bandits, Optimization, Probabilistic Graphical Models, Learning from Graph-structured Data/Graph Recovery, Learning / Making Predictions from Noisy Data, Python, PyTorch, Matlab, Scikit-learnProfessional ExperienceYale University New Haven, ConnecticutPOSTDOCTORAL ASSOCIATE, SCHOOLOF ENGINEERING & APPLIED SCIENCE, ELECTRICAL ENGINEERING Jul 2021 - Present Research Domain: Machine learning, Algorithmic generalization error analysis and robustness in learningMajor Contributions: Developed novel statistical guarantees for accuracy in learning; Generalization error bounds for optimization algorithms including Stochastic Gradient Descent (SGD), variations of (S)GD, and Zeroth-Order learning (Derivative Free Optimization). Designed a novel robust federated learning algorithm under restricted user availability. Introduced (S)GD variants for minimizing time to accuracy for deep neural networks training. Established novel statistical (sample complexity) guarantees for quantile-multi armed bandits with applications to A/B testing. Designed the first algorithm for quantile A/B testing under differential privacy. Proposed the first statistical analysis for learning graphs through structured data under the presence of noise. Established novel statistical guarantees for accurate prediction making and inference on noisy tree-structured data, with applications on recommender systems, epidemic and population dynamics. Detailed information of my research and major contributions appear in Publications below (and Google Scholar). Michigan State University East Lansing, MichiganPOSTDOCTORAL RESEARCHER, DEPARTMENT OF ELECTRICAL AND COMPUTER ENGINEERING May 2021 - Jun 2021 Host: Prof. Dionysios Kalogerias Core Project: Multi-agent decision making approaches with applications on supply chain optimization Inventorymanagement; designing dynamic input-output PDE - ODE models for supply chain applications. EducationRutgers University, Electrical and Computer Engineering New Brunswick, New Jersey PH.D. IN ELECTRICAL AND COMPUTER ENGINEERING Feb 2016 - Apr 2021 Thesis: Learning Tree-Structured Models from Noisy Data Adviser: Prof. Anand Sarwate Selected Courses: Error Control Coding, Information Theory, Measure Theory, Convex Optimization GPA: 3.9/4Rutgers University, Computer Science New Brunswick, New Jersey PH.D. CANDIDATE IN COMPUTER SCIENCE Sep 2014 -Jan 2016 Major: Machine Learning and Pattern Recognition Adviser: Prof. Dimitris Metaxas Selected Courses: Stochastic Dynamic Programming, Advanced Algorithms, Pattern Recognition, Numerical Analysis GPA: 3.8/4University of Patras Patras, GreeceDIPLOMA OF ELECTRICAL AND COMPUTER ENGINEERING, BACHELORS AND MASTERS DEGREE Sep 2009 -Jun 2014 Major: Stochastic Signal Processing and Communications Adviser: Prof. George Moustakides GPA: 8.3/10, top 4% in class of 2013-2014 (250 students, ECE Department) Programming Competition: 1st place (among 250 participants, class of 2013-2014) MAY 24, 2024 KONSTANTINOS NIKOLAKAKIS  RSUMPublications K. Nikolakakis, G. Chantzialexiou, D. Kalogerias, FEDSTR: Money-In AI-Out A Decentralized Marketplace for Federated Learning and LLM Training on the NOSTR Protocol, Preprint K. Nikolakakis, A. Karbasi, D. Kalogerias, Select without Fear: Almost All Mini-Batch Schedules Generalize Optimally, Under review P. Theodoropoulos, K. Nikolakakis, D. Kalogerias, Federated Learning Under Restricted User Availability, IEEE International Confer- ence on Acoustics, Speech and Signal Processing, (ICASSP 2024) P. Okanovic, R. Waleffe, V. Mageirakos, K. Nikolakakis et al., RepeatedRandomSamplingforMinimizingtheTime-to-Accuracy of Learn- ing, International Conference on Learning Representations, (ICLR 2024) K. Nikolakakis, F. Haddadpour, A. Karbasi, D. Kalogerias, BeyondLipschitz:SharpGeneralizationandExcessRiskBoundsforFull-Batch Gradient Descent, International Conference on Learning Representations, ICLR 2023 K. Nikolakakis, F. Haddadpour, D. Kalogerias, A. Karbasi, Black-Box Generalization: Stability of Zeroth-Order Learning, Neural Infor- mation Processing Systems, (NeurIPS 2022) K. Nikolakakis, D. Kalogerias, A. Sarwate, Optimal Rates for Learning Hidden Tree Structures, ArXiv Preprint K.Nikolakakis, D. Kalogerias, O. Sheffet,A. Sarwate,QuantileMulti-ArmedBandits:OptimalBest-ArmIdentificationandaDifferentially Private Scheme, IEEE Journal on Selected Areas in Information Theory, (JSAIT), May 2021 K. Nikolakakis, D. Kalogerias, A. Sarwate, Predictive Learning on Hidden Tree-Structured Ising Models, Journal of Machine Learning Research (JMLR 2021), vol. 22, no 59, pp. 1-82, February 2021 K. Nikolakakis, D. Kalogerias, A. Sarwate, Learning Tree Structures from Noisy Data, 22nd International Conference on Artificial In- telligence and Statistics (AISTATS 2019)Invited Talks Annual Conference on Information Science and Systems (CISS), Princeton University, Princeton, New Jersey, 2020 Conference on Inference on Graphical Models, Columbia University, New York, 2019 UC San Diego (Prof. Mikhail Belkins group), Beyond Lipschitz: Sharp Generalization and Excess Risk Bounds for Full-Batch Gradient DescentAwards Anthony Massini Postdoctoral Fellowship: One year of support for postdoctoral studies, Yale University, 2022-2023 NeurIPS 2022 Scholar Award, Travel scholarship, NeurIPS 2022 Gerondelis Scholarship by Gerondelis Foundation INC., Rutgers University, 2016-2017 Programming Skills Python, PyTorch, Scikit-learn, Matlab, C, C++, Unix Teaching ExperienceRutgers University New Brunswick, New JerseyDEPARTMENTS OF ELECTRICAL AND COMPUTER ENGINEERING & COMPUTER SCIENCE Sep 2014 - Jan 2017 Probability Theory and Stochastic Processes (Spring 2016, Fall 2016, Fall 2017) Programming in C & Unix (Fall 2014, Spring 2015, Fall 2015) Project Management Skills Goal setting and monitoring skills: Proposals for research funding (Anthony Massini Award), Course/Teaching Planning People management skills: Mentoring and guidance of Ph.D. students Teamwork: Collaboration with researchers from Yale and other institutions including ETH (Zurich), Princeton, Wisconsin-Madison Leading experience, initiatives and risk management; Commitment of resources, Identification of open research challenges Languages English, GreekHobbies Martial Arts, Photography, SnowboardingMAY 24, 2024 KONSTANTINOS NIKOLAKAKIS  RSUM