As a pervasive issue, missing data may influence the data modeling performance and lead to more difficulties of completing the desired tasks. Many approaches have been developed for missing data imputation. Recently, by taking advantage of the emerging generative adversarial network (GAN), an effective missing data imputation approach termed generative adversarial imputation nets (GAIN) was developed. However, its modeling architecture may still lead to significant imputation bias. In addition, with the GAN structure, the training process of GAIN may be unstable and the imputation variation may be high. Hence, to address these two limitations, the ensemble GAIN with selective multi-generator (ESM-GAIN) is proposed to improve the imputation accuracy and robustness. The contributions of the proposed ESM-GAIN consist of two aspects: (1) a selective multi-generation framework is proposed to identify high-quality imputations; (2) an ensemble learning framework is incorporated for GAIN imputation to improve the imputation robustness. The effectiveness of the proposed ESM-GAIN is validated by both numerical simulation and two real-world breast cancer datasets. 
                        more » 
                        « less   
                    
                            
                            Transfer learning-based ROADM EDFA wavelength dependent gain prediction using minimized data collection
                        
                    
    
            We implement and test transfer learning-based gain models across 16 ROADM EDFAs, which achieve less than 0.17/0.30 dB mean absolute error for booster/pre-amplifier gain prediction using only 0.5% of the full target EDFA dataset. 
        more » 
        « less   
        
    
                            - Award ID(s):
- 2029295
- PAR ID:
- 10457282
- Date Published:
- Journal Name:
- in Proc. IEEE/OPTICA Optical Fiber Communication Conference (OFC’23), Th2A.1, Mar. 2023
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
- 
            
- 
            The self-fitting Bose SoundControl™ hearing aid is the first of its kind to gain FDA clearance. In the self-fitting process, the Bose Hear app uses the Bose CustomTune™ interface for mapping to a wide range of target gain profiles, derived from a hearing loss database. This article compares the population coverage—or the percentage of people who would be able to find a frequency gain profile similar to a NAL-NL2 prescription fit—of SoundControl to other self-fitting amplification devices which typically feature only 1 to 4 preset gain profiles.more » « less
- 
            Colloidal semiconductor nanocrystals (NCs) have emerged as promising candidates for developing solutionprocessable optical gain media with potential applications in integrated photonic circuits and lasers. However, the deployment of NCs in these technologies has been hindered by the nonradiative Auger recombination of multiexciton states, which shortens the optical gain lifetime and reduces its spectral range. Here, we demonstrate that these limitations can be overcome by using giant colloidal quantum shells (g-QSs), comprising a quantum-confined CdSe shell grown over a large (∼14 nm) CdS bulk core. Such bulk-nanoscale architecture minimizes exciton− exciton interactions, leading to suppressed Auger recombination and one of the broadest gain bandwidths reported for colloidal nanomaterials, spanning energies both above and, remarkably, below the bandgap. Ultrafast transient absorption and photoluminescence measurements demonstrate that the high-energy portion of optical gain arises from states containing more than 15 excitons per particle, while the unusual sub-bandgap gain behavior results from an Auger-assisted radiative recombination, a mechanism that has traditionally been viewed as a loss pathway. Collectively, these results reveal a unique gain regime associated with bulk-nanocrystal hybrid systems, which offers a promising path toward solution-processable light sources.more » « less
- 
            A linear time-invariant (LTI) system is usually characterized by its gains, which are ratios of various output and input signal magnitudes. When the main function of the system is energy conversion, one is more interested in the power or energy gain or loss between input and output terminals. Such a "gain" represents the efficiency of the system as an energy conversion device. In this paper, we show that the power or energy "gain" of an LTI system can be determined as a nonlinear function of three signal gains of the system. This results in an explicit formula for the power (energy) gain in terms of the system parameters. This can then be optimized over the design parameters to maximize the gain and thus the system energy efficiency. A detailed DC circuit example is included for illustration. The energy efficiency of an LTI system operating in a steady state in response to a sinusoidal signal is also determined.more » « less
- 
            The bag gain relates to a gain in security due to spreading payload among multiple covers when the steganog- rapher maintains a positive communication rate. This gain is maximal for a certain optimal bag size, which depends on the embedding method, payload spreading strategy, communication rate, and the cover source. Originally discovered and analyzed in the spatial domain, in this paper we study this phenomenon for JPEG images across quality factors. Our experiments and theoretical analysis indicate that the bag gain is more pronounced for higher JPEG qualities, more aggressive batch senders, and for senders maintaining a fixed payload per bag in terms of bits per DCT rather than per non-zero AC DCT.more » « less
 An official website of the United States government
An official website of the United States government 
				
			 
					 
					
 
                                    