1
|
Ahangaran M, Zhu H, Li R, Yin L, Jang J, Chaudhry AP, Farrer LA, Au R, Kolachalama VB. DREAMER: a computational framework to evaluate readiness of datasets for machine learning. BMC Med Inform Decis Mak 2024; 24:152. [PMID: 38831432 PMCID: PMC11149315 DOI: 10.1186/s12911-024-02544-w] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/13/2023] [Accepted: 05/20/2024] [Indexed: 06/05/2024] Open
Abstract
BACKGROUND Machine learning (ML) has emerged as the predominant computational paradigm for analyzing large-scale datasets across diverse domains. The assessment of dataset quality stands as a pivotal precursor to the successful deployment of ML models. In this study, we introduce DREAMER (Data REAdiness for MachinE learning Research), an algorithmic framework leveraging supervised and unsupervised machine learning techniques to autonomously evaluate the suitability of tabular datasets for ML model development. DREAMER is openly accessible as a tool on GitHub and Docker, facilitating its adoption and further refinement within the research community.. RESULTS The proposed model in this study was applied to three distinct tabular datasets, resulting in notable enhancements in their quality with respect to readiness for ML tasks, as assessed through established data quality metrics. Our findings demonstrate the efficacy of the framework in substantially augmenting the original dataset quality, achieved through the elimination of extraneous features and rows. This refinement yielded improved accuracy across both supervised and unsupervised learning methodologies. CONCLUSION Our software presents an automated framework for data readiness, aimed at enhancing the integrity of raw datasets to facilitate robust utilization within ML pipelines. Through our proposed framework, we streamline the original dataset, resulting in enhanced accuracy and efficiency within the associated ML algorithms.
Collapse
Affiliation(s)
- Meysam Ahangaran
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Hanzhi Zhu
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Ruihui Li
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Lingkai Yin
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Joseph Jang
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Arnav P Chaudhry
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Lindsay A Farrer
- Department of Neurology, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
- Department Ophthalmology, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
- Department of Epidemiology, Boston University School of Public Health, Boston, MA, USA
- Department of Biostatistics, Boston University School of Public Health, Boston, MA, USA
- Boston University Alzheimer's Disease Research Center, Boston, MA, USA
- The Framingham Heart Study, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Rhoda Au
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
- Department of Neurology, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
- Department of Epidemiology, Boston University School of Public Health, Boston, MA, USA
- Boston University Alzheimer's Disease Research Center, Boston, MA, USA
- The Framingham Heart Study, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
- Department of Anatomy and Neurobiology, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA
| | - Vijaya B Kolachalama
- Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, MA, USA.
- Boston University Alzheimer's Disease Research Center, Boston, MA, USA.
- Department of Computer Science, Boston University, Boston, MA, USA.
- Faculty of Computing & Data Sciences, Boston University, Boston, MA, 02215, USA.
| |
Collapse
|
2
|
Li XK, Ma JX, Li XY, Hu JJ, Ding CY, Han FK, Guo XM, Tan X, Jin XM. High-efficiency reinforcement learning with hybrid architecture photonic integrated circuit. Nat Commun 2024; 15:1044. [PMID: 38316815 PMCID: PMC10844654 DOI: 10.1038/s41467-024-45305-z] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/09/2023] [Accepted: 01/18/2024] [Indexed: 02/07/2024] Open
Abstract
Reinforcement learning (RL) stands as one of the three fundamental paradigms within machine learning and has made a substantial leap to build general-purpose learning systems. However, using traditional electrical computers to simulate agent-environment interactions in RL models consumes tremendous computing resources, posing a significant challenge to the efficiency of RL. Here, we propose a universal framework that utilizes a photonic integrated circuit (PIC) to simulate the interactions in RL for improving the algorithm efficiency. High parallelism and precision on-chip optical interaction calculations are implemented with the assistance of link calibration in the hybrid architecture PIC. By introducing similarity information into the reward function of the RL model, PIC-RL successfully accomplishes perovskite materials synthesis task within a 3472-dimensional state space, resulting in a notable 56% improvement in efficiency. Our results validate the effectiveness of simulating RL algorithm interactions on the PIC platform, highlighting its potential to boost computing power in large-scale and sophisticated RL tasks.
Collapse
Affiliation(s)
- Xuan-Kun Li
- Center for Integrated Quantum Information Technologies (IQIT), School of Physics and Astronomy and State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, 200240, China
- Hefei National Laboratory, Hefei, 230088, China
| | - Jian-Xu Ma
- TuringQ Co., Ltd., Shanghai, 200240, China
| | | | - Jun-Jie Hu
- Center for Integrated Quantum Information Technologies (IQIT), School of Physics and Astronomy and State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, 200240, China
- Hefei National Laboratory, Hefei, 230088, China
| | - Chuan-Yang Ding
- Center for Integrated Quantum Information Technologies (IQIT), School of Physics and Astronomy and State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, 200240, China
- Hefei National Laboratory, Hefei, 230088, China
| | - Feng-Kai Han
- Center for Integrated Quantum Information Technologies (IQIT), School of Physics and Astronomy and State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, 200240, China
- Hefei National Laboratory, Hefei, 230088, China
| | | | - Xi Tan
- Center for Integrated Quantum Information Technologies (IQIT), School of Physics and Astronomy and State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, 200240, China
- Hefei National Laboratory, Hefei, 230088, China
| | - Xian-Min Jin
- Center for Integrated Quantum Information Technologies (IQIT), School of Physics and Astronomy and State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, 200240, China.
- Hefei National Laboratory, Hefei, 230088, China.
- TuringQ Co., Ltd., Shanghai, 200240, China.
- Chip Hub for Integrated Photonics Xplore (CHIPX), Shanghai Jiao Tong University, Wuxi, 214000, China.
| |
Collapse
|
3
|
Wu C, Deng H, Huang YS, Yu H, Takeuchi I, Ríos Ocampo CA, Li M. Freeform direct-write and rewritable photonic integrated circuits in phase-change thin films. SCIENCE ADVANCES 2024; 10:eadk1361. [PMID: 38181081 PMCID: PMC10775994 DOI: 10.1126/sciadv.adk1361] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 08/04/2023] [Accepted: 12/01/2023] [Indexed: 01/07/2024]
Abstract
Photonic integrated circuits (PICs) with rapid prototyping and reprogramming capabilities promise revolutionary impacts on a plethora of photonic technologies. We report direct-write and rewritable photonic circuits on a low-loss phase-change material (PCM) thin film. Complete end-to-end PICs are directly laser-written in one step without additional fabrication processes, and any part of the circuit can be erased and rewritten, facilitating rapid design modification. We demonstrate the versatility of this technique for diverse applications, including an optical interconnect fabric for reconfigurable networking, a photonic crossbar array for optical computing, and a tunable optical filter for optical signal processing. By combining the programmability of the direct laser writing technique with PCM, our technique unlocks opportunities for programmable photonic networking, computing, and signal processing. Moreover, the rewritable photonic circuits enable rapid prototyping and testing in a convenient and cost-efficient manner, eliminate the need for nanofabrication facilities, and thus promote the proliferation of photonics research and education to a broader community.
Collapse
Affiliation(s)
- Changming Wu
- Department of Electrical and Computer Engineering, University of Washington, Seattle, WA 98195, USA
| | - Haoqin Deng
- Department of Electrical and Computer Engineering, University of Washington, Seattle, WA 98195, USA
| | - Yi-Siou Huang
- Department of Materials Science and Engineering, University of Maryland, College Park, MD 20742, USA
- Institute for Research in Electronics and Applied Physics, University of Maryland, College Park, MD 20742, USA
| | - Heshan Yu
- Department of Materials Science and Engineering, University of Maryland, College Park, MD 20742, USA
- School of Microelectronics, Tianjin University, Tianjin 300072, China
| | - Ichiro Takeuchi
- Department of Materials Science and Engineering, University of Maryland, College Park, MD 20742, USA
| | - Carlos A. Ríos Ocampo
- Department of Materials Science and Engineering, University of Maryland, College Park, MD 20742, USA
- Institute for Research in Electronics and Applied Physics, University of Maryland, College Park, MD 20742, USA
| | - Mo Li
- Department of Electrical and Computer Engineering, University of Washington, Seattle, WA 98195, USA
- Department of Physics, University of Washington, Seattle, WA 98195, USA
| |
Collapse
|
4
|
Kaushal S, Aadhi A, Roberge A, Morandotti R, Kashyap R, Azaña J. All-fibre phase filters with 1-GHz resolution for high-speed passive optical logic processing. Nat Commun 2023; 14:1808. [PMID: 37002203 PMCID: PMC10066316 DOI: 10.1038/s41467-023-37472-2] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/28/2022] [Accepted: 03/10/2023] [Indexed: 04/03/2023] Open
Abstract
Photonic-based implementation of advanced computing tasks is a potential alternative to mitigate the bandwidth limitations of electronics. Despite the inherent advantage of a large bandwidth, photonic systems are generally bulky and power-hungry. In this respect, all-pass spectral phase filters enable simultaneous ultrahigh speed operation and minimal power consumption for a wide range of signal processing functionalities. Yet, phase filters offering GHz to sub-GHz frequency resolution in practical, integrated platforms have remained elusive. We report a fibre Bragg grating-based phase filter with a record frequency resolution of 1 GHz, at least 10× improvement compared to a conventional optical waveshaper. The all-fibre phase filter is employed to experimentally realize high-speed fully passive NOT and XNOR logic operations. We demonstrate inversion of a 45-Gbps 127-bit random sequence with an energy consumption of ~34 fJ/bit, and XNOR logic at a bit rate of 10.25 Gbps consuming ~425 fJ/bit. The scalable implementation of phase filters provides a promising path towards widespread deployment of compact, low-energy-consuming signal processors.
Collapse
Affiliation(s)
- Saket Kaushal
- Énergie, Matériaux et Télécommunications, Institut National de la Recherche Scientifique, 1650 Lionel-Boulet Blvd., Varennes, J3X 1P7, Quebec, Canada
| | - A Aadhi
- Énergie, Matériaux et Télécommunications, Institut National de la Recherche Scientifique, 1650 Lionel-Boulet Blvd., Varennes, J3X 1P7, Quebec, Canada
| | - Anthony Roberge
- Department of Engineering Physics, Fabulas Laboratory, Polytechnique Montréal, 2500 Chem. de Polytechnique, Montréal, H3T 1J4, Quebec, Canada
| | - Roberto Morandotti
- Énergie, Matériaux et Télécommunications, Institut National de la Recherche Scientifique, 1650 Lionel-Boulet Blvd., Varennes, J3X 1P7, Quebec, Canada
| | - Raman Kashyap
- Department of Engineering Physics, Fabulas Laboratory, Polytechnique Montréal, 2500 Chem. de Polytechnique, Montréal, H3T 1J4, Quebec, Canada
- Department of Electrical Engineering, Fabulas Laboratory, Polytechnique Montréal, 2500 Chem. de Polytechnique, Montréal, H3T 1J4, Quebec, Canada
| | - José Azaña
- Énergie, Matériaux et Télécommunications, Institut National de la Recherche Scientifique, 1650 Lionel-Boulet Blvd., Varennes, J3X 1P7, Quebec, Canada.
| |
Collapse
|
5
|
Towards a high-density photonic tensor core enabled by intensity-modulated microrings and photonic wire bonding. Sci Rep 2023; 13:1260. [PMID: 36690656 PMCID: PMC9870901 DOI: 10.1038/s41598-023-27724-y] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/26/2022] [Accepted: 01/06/2023] [Indexed: 01/24/2023] Open
Abstract
We propose a photonic processing unit for high-density analog computation using intensity-modulation-based microring modulators (IM-MRMs). The output signal at the fixed resonance wavelength is directly intensity modulated by changing the extinction ratio (ER) of the IM-MRM . Thanks to the intensity-modulated approach, the proposed photonic processing unit is less sensitive to the inter-channel crosstalk. Simulation results reveal that the proposed design offers a maximum of 17-fold increase in wavelength channel density compared to its wavelength-modulated counterpart. Therefore, a photonic tensor core of size 512 [Formula: see text] 512 can be realized by current foundry lines. A convolutional neural network (CNN) simulator with 6-bit precision is built for handwritten digit recognition task using the proposed modulator. Simulation results show an overall accuracy of 96.76%, when the wavelength channel spacing suffers a 3-dB power penalty. To experimentally validate the system, 1000 dot product operations are carried out with a 4-bit signed system on a co-packaged photonic chip, where optical and electrical I/Os are realized using photonic and electrical wire bonding techniques. Study of the measurement results show a mean squared error (MSE) of 3.09[Formula: see text]10[Formula: see text] for dot product calculations. The proposed IM-MRM, therefore, renders the crosstalk issue tractable and provides a solution for the development of large-scale optical information processing systems with multiple wavelengths.
Collapse
|
6
|
Zhu H, Wu B, Gao M, Ren F, Qin W, Juodkazis S, Chen F. Femtosecond Laser Direct‐Write Plasmonic Nanolithography in Dielectrics. SMALL SCIENCE 2022. [DOI: 10.1002/smsc.202200038] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/11/2022] Open
Affiliation(s)
- Han Zhu
- School of Physics State Key Laboratory of Crystal Materials Shandong University Jinan 250100 China
| | - Bo Wu
- School of Physics State Key Laboratory of Crystal Materials Shandong University Jinan 250100 China
| | - Mingsheng Gao
- School of Physics State Key Laboratory of Crystal Materials Shandong University Jinan 250100 China
| | - Feng Ren
- Department of Physics Center for Ion Beam Application and Center for Electron Microscopy Wuhan University Wuhan 430072 China
| | - Wei Qin
- School of Physics State Key Laboratory of Crystal Materials Shandong University Jinan 250100 China
| | - Saulius Juodkazis
- Optical Sciences Centre Faculty of Science, Engineering and Technology Swinburne University of Technology Hawthorn VIC 3122 Australia
| | - Feng Chen
- School of Physics State Key Laboratory of Crystal Materials Shandong University Jinan 250100 China
| |
Collapse
|
7
|
Henglein F, Shoham S, Vizel Y. Formal Semantics and Verification of Network-Based Biocomputation Circuits. LECTURE NOTES IN COMPUTER SCIENCE 2021. [PMCID: PMC7798404 DOI: 10.1007/978-3-030-67067-2_21] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Indexed: 11/25/2022]
Abstract
Network-Based Biocomputation Circuits (NBCs) offer a new paradigm for solving complex computational problems by utilizing biological agents that operate in parallel to explore manufactured planar devices. The approach can also have future applications in diagnostics and medicine by combining NBCs computational power with the ability to interface with biological material. To realize this potential, devices should be designed in a way that ensures their correctness and robust operation. For this purpose, formal methods and tools can offer significant advantages by allowing investigation of design limitations and detection of errors before manufacturing and experimentation. Here we define a computational model for NBCs by providing formal semantics to NBC circuits. We present a formal verification-based approach and prototype tool that can assist in the design of NBCs by enabling verification of a given design’s correctness. Our tool allows verification of the correctness of NBC designs for several NP-Complete problems, including the Subset Sum, Exact Cover and Satisfiability problems and can be extended to other NBC implementations. Our approach is based on defining transition systems for NBCs and using temporal logic for specifying and proving properties of the design using model checking. Our formal model can also serve as a starting point for computational complexity studies of the power and limitations of NBC systems.
Collapse
|
8
|
Li M, Zhang Q, Chen Y, Ren X, Gong Q, Li Y. Femtosecond Laser Direct Writing of Integrated Photonic Quantum Chips for Generating Path-Encoded Bell States. MICROMACHINES 2020; 11:mi11121111. [PMID: 33334077 PMCID: PMC7765531 DOI: 10.3390/mi11121111] [Citation(s) in RCA: 6] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 11/27/2020] [Revised: 12/12/2020] [Accepted: 12/13/2020] [Indexed: 11/16/2022]
Abstract
Integrated photonic quantum chip provides a promising platform to perform quantum computation, quantum simulation, quantum metrology and quantum communication. Femtosecond laser direct writing (FLDW) is a potential technique to fabricate various integrated photonic quantum chips in glass. Several quantum logic gates fabricated by FLDW have been reported, such as polarization and path encoded quantum controlled-NOT (CNOT) gates. By combining several single qubit gates and two qubit gates, the quantum circuit can realize different functions, such as generating quantum entangled states and performing quantum computation algorithms. Here we demonstrate the FLDW of integrated photonic quantum chips composed of one Hadamard gate and one CNOT gate for generating all four path-encoded Bell states. The experimental results show that the average fidelity of the reconstructed truth table reaches as high as 98.8 ± 0.3%. Our work is of great importance to be widely applied in many quantum circuits, therefore this technique would offer great potential to fabricate more complex circuits to realize more advanced functions.
Collapse
Affiliation(s)
- Meng Li
- State Key Laboratory for Mesoscopic Physics and Frontiers Science Center for Nano-Optoelectronics, School of Physics, Peking University, Beijing 100871, China; (M.L.); (Q.Z.); (Q.G.)
| | - Qian Zhang
- State Key Laboratory for Mesoscopic Physics and Frontiers Science Center for Nano-Optoelectronics, School of Physics, Peking University, Beijing 100871, China; (M.L.); (Q.Z.); (Q.G.)
| | - Yang Chen
- CAS Key Laboratory of Quantum Information, University of Science and Technology of China, Hefei 230026, China; (Y.C.); (X.R.)
| | - Xifeng Ren
- CAS Key Laboratory of Quantum Information, University of Science and Technology of China, Hefei 230026, China; (Y.C.); (X.R.)
| | - Qihuang Gong
- State Key Laboratory for Mesoscopic Physics and Frontiers Science Center for Nano-Optoelectronics, School of Physics, Peking University, Beijing 100871, China; (M.L.); (Q.Z.); (Q.G.)
- Collaborative Innovation Center of Extreme Optics, Shanxi University, Taiyuan 030006, China
- Peking University Yangtze Delta Institute of Optoelectronics, Nantong 226010, China
| | - Yan Li
- State Key Laboratory for Mesoscopic Physics and Frontiers Science Center for Nano-Optoelectronics, School of Physics, Peking University, Beijing 100871, China; (M.L.); (Q.Z.); (Q.G.)
- Collaborative Innovation Center of Extreme Optics, Shanxi University, Taiyuan 030006, China
- Peking University Yangtze Delta Institute of Optoelectronics, Nantong 226010, China
- Correspondence:
| |
Collapse
|
9
|
Ryabov A, Thurner JW, Nabben D, Tsarev MV, Baum P. Attosecond metrology in a continuous-beam transmission electron microscope. SCIENCE ADVANCES 2020; 6:6/46/eabb1393. [PMID: 33177078 PMCID: PMC7673721 DOI: 10.1126/sciadv.abb1393] [Citation(s) in RCA: 21] [Impact Index Per Article: 5.3] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 02/02/2020] [Accepted: 09/25/2020] [Indexed: 05/12/2023]
Abstract
Electron microscopy can visualize the structure of complex materials with atomic and subatomic resolution, but investigations of reaction dynamics and light-matter interaction call for time resolution as well, ideally on a level below the oscillation period of light. Here, we report the use of the optical cycles of a continuous-wave laser to bunch the electron beam inside a transmission electron microscope into electron pulses that are shorter than half a cycle of light. The pulses arrive at the target at almost the full average brightness of the electron source and in synchrony to the optical cycles, providing attosecond time resolution of spectroscopic features. The necessary modifications are simple and can turn almost any electron microscope into an attosecond instrument that may be useful for visualizing the inner workings of light-matter interaction on the basis of the atoms and the cycles of light.
Collapse
Affiliation(s)
- A Ryabov
- Universität Konstanz, Universitätsstr. 10, 78464 Konstanz, Germany
- Ludwig-Maximilians-Universität München, Am Coulombwall 1, 85748 Garching, Germany
| | - J W Thurner
- Ludwig-Maximilians-Universität München, Am Coulombwall 1, 85748 Garching, Germany
| | - D Nabben
- Universität Konstanz, Universitätsstr. 10, 78464 Konstanz, Germany
| | - M V Tsarev
- Ludwig-Maximilians-Universität München, Am Coulombwall 1, 85748 Garching, Germany.
| | - P Baum
- Universität Konstanz, Universitätsstr. 10, 78464 Konstanz, Germany.
- Ludwig-Maximilians-Universität München, Am Coulombwall 1, 85748 Garching, Germany
| |
Collapse
|