Martinez-Villaseñor, Lourdes
Thumbnail Image
Preferred name
Martinez-Villaseñor, Lourdes
Official Name
Martínez Villaseñor, María de Lourdes
Alternative Name
lmartine
Main Affiliation
ORCID
Scopus Author ID
55521085100
Researcher ID
N-7607-2018

Research Output
Now showing 1 - 3 of 3
No Thumbnail Available
Publication

Challenges in Data Acquisition Systems: Lessons Learned from Fall Detection to Nanosensors

2018 , Peñafort Asturiano, Carlos J. , Santiago, Nestor , Nuñez-Martínez, José , Ponce, Hiram , Martinez-Villaseñor, Lourdes

Falls are a major public health problem in elderly people often causing fatal injuries. It is important to assure that injured people receive assistance as quick as possible. Fall detection systems have gain more relevance nowadays. As more databases and fall detection systems are developed, there is more need to identify the challenges encountered in building and creating them. This paper addresses pre-processing, inconsistency and synchronization challenges that occur when creating a multimodal database for fall detection. We present different algorithms used to tackle these issues. We describe the issues and the corresponding solutions in order to document the lessons learned that could help others in data acquisition for multimodal databases. Applying the solutions to the issues found so far, we acquired an organized multimodal database for fall detection with 17 subjects. Furthermore, these lessons learned can be applied for data nanosensors acquisition and storage. © 2018 IEEE.

View
No Thumbnail Available
Publication

Comparative Analysis of Artificial Hydrocarbon Networks versus Convolutional Neural Networks in Human Activity Recognition

2020 , Ponce, Hiram , Martinez-Villaseñor, Lourdes

Human activity recognition (HAR) has gained interest in the research communities in order to know the behavior and context of users for medical, sports performance evaluation, ambient assisted living and security applications. Recent works suggest that convolutional neural networks (CNN) are very competitive machine learning techniques for HAR. Nevertheless, CNN require many computational resources, high number of parameter tuning, and many data samples for training. In this paper, we present a comparative analysis of a novel technique, artificial hydrocarbon networks (AHN), with CNN on human activity recognition classification task. We choose to compare AHN with CNN given that it is a very well-suited machine learning technique for HAR. We show that AHN architecture is simpler to set up than CNN, it needs less hyper-parameter configuration and has a slightly better accuracy performance. © 2020 IEEE.

View
No Thumbnail Available
Publication

Feature Selection Methods Evaluation for CTR Estimation

2016 , Miralles-Pechuán, Luis , Ponce, Hiram , Martinez-Villaseñor, Lourdes

The most widespread payment model in online advertising is Cost-per-click (CPC). In this model the advertisers pay each time that a user generates a click. In order to enhance the income of CPC Advertising Networks, it is necessary to give priority to the most profitable adverts. The most important factor in the profitability of an advert is Click-through-rate (CTR), which is the probability that a user generates a click in a given advert. In this paper we find which feature selection method between PCA, RFE, Gain ratio and NSGA-II is better suited, or if otherwise, the machine learning classification methods work best without any feature selection method. ©2016 IEE

View