Daniel Lopez-Martinez

Daniel Lopez-Martinez

Auto-generated profile, based on publicly available information

Daniel Lopez-Martinez

Instituto Politécnico Nacional

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 October 2019 in IEEE Transactions on Affective Computing
Verified
Journal Article
Closed access
PWPhilipp Werner
Philipp Werner
DLDaniel Lopez-Martinez
Daniel Lopez-Martinez
SWSteffen Walter
Steffen Walter
AAAyoub Al-Hamadi
Ayoub Al-Hamadi
SGSascha Gruss
Sascha Gruss
RPRosalind W. Picard
Rosalind W. Picard

1 January 2021 in Agronomía Mesoamericana
Verified
Journal Article
Open access
PDF file
JAJosé Luis Arellano-Vázquez
José Luis Arellano-Vázquez
GGGermán Fernando Gutiérrez-Hernández
Germán Fernando Gutiérrez-Hernández
EFEstela Flores-Gómez
Estela Flores-Gómez
DLDaniel Eduardo López-Martínez
Daniel Eduardo López-Martínez
Introduction. Currently, maize is the most cultivated grain in the world and it is feasible to improve the nutrition of its consumers through the development of quality protein maize (QPM), whose basis is the expression of the opaque2 gene, which increases the contents of lysine and tryptophan in the grain. Objective. To analyze chemically the protein of the grain of lines and experimental simple hybrids of QPM and to know the relationships between the genotype and the evaluated variables of protein quality. Materials and methods. The experiments were carried out during 2017 in the Interdisciplinary Professional Unit of Biotechnology of the Instituto Politecnico Nacional (Mexico City). Whole grains of parental lines (P1 to P6) and direct simple hybrids were used. Lysine, tryptophan, and protein quantity and quality were estimated. A completely randomized design with three replications was used. Analysis of variance, comparison of means, and principal component analysis were performed. Results. Lines P2, P3, and P6, and hybrids P2 x P4 and P2 x P6 were identified as significant (p ² 0.05) QPM. The protein quality index was correlated with tryptophan (r = 0.9 *) and with lysine (r = 0.7 *) but not with total protein. Conclusions. The fundamental factor for developing the QPM were the essential amino acids and the limiting factor for increasing their nutritional value was the quality of the protein.

20 November 2022 in arXiv
Verified
Preprint
Open access
PDF file
DLDaniel Lopez-Martinez
Daniel Lopez-Martinez
AYAlex Yakubovich
Alex Yakubovich
MSMartin Seneviratne
Martin Seneviratne
ALAdam D. Lelkes
Adam D. Lelkes
ATAkshit Tyagi
Akshit Tyagi
Mingjun ChenMingjun Chen
Mingjun Chen
While it has been well known in the ML community that deep learning models suffer from instability, the consequences for healthcare deployments are under characterised. We study the stability of different model architectures trained on electronic health records, using a set of outpatient prediction tasks as a case study. We show that repeated training runs of the same deep learning model on the same training data can result in significantly different outcomes at a patient level even though global performance metrics remain stable. We propose two stability metrics for measuring the effect of randomness of model training, as well as mitigation strategies for improving model stability.