{"title":"Impacts, Potentials, and Trends of Machine Learning in Industry 4.0: An Overview","authors":"Nuvia Da Silva, M. Oliveira, D. Cordeiro","doi":"10.52591/lxai202007136","DOIUrl":"https://doi.org/10.52591/lxai202007136","url":null,"abstract":"Intelligent techniques has been required at the Industry 4.0 to integrate resources, data and technology. Many manufacturing problems can use machine learning to improve the service or product by analyzing, learning and correlating data stored in large databases. Thus, machine learning is essential to predict the process lead time, which is a fundamental parameter for the success of management and production planning and control. In addition, there is a range of machine learning applications in manufacturing problems that have not yet been fully investigated.","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123904256","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Topologically-based Variational Autoencoder for Time Series Classification","authors":"Rodrigo Rivera","doi":"10.52591/lxai202007139","DOIUrl":"https://doi.org/10.52591/lxai202007139","url":null,"abstract":"Topological Data Analysis is an approach to analyze data using different techniques from topology. These techniques aim to extract fundamental qualitative properties, such as shape and connectivity in data. In this work, we propose a universal approach for time series classification with variational autoencoders. It is built on extracted features from the persistent homology theory. Compared to standard classification approaches, the proposed methodology enables the classification of time series, which have different recurrent behavior in the reconstructed phase space. Multiple experiments with time-series datasets confirm that the method makes classification more robust to noisy and high-dimensional data and favors datasets in which shape has meaning.","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124100026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A Bayesian time series model of Coca leaf production in Colombia","authors":"Segio Garrido, Emiliano Isaza","doi":"10.52591/lxai202007132","DOIUrl":"https://doi.org/10.52591/lxai202007132","url":null,"abstract":"We propose a Bayesian model to predict Coca production in Colombia so that policy makers can take uncertainty into account in their decision making process. Given the high dimensionality of the problem and the relatively low quantity of data points, we propose a Hierarchical time series approach. Even though our model performs better compared to three other specifications using Leave Future Out Cross Validation, some future modeling approaches are suggested to improve the model’s performance.","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"117 15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126402127","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Laguerre-Gauss Preprocessing: Line Profiles as Image Features for Aerial Images Classification","authors":"Alejandro Murillo","doi":"10.52591/lxai202007131","DOIUrl":"https://doi.org/10.52591/lxai202007131","url":null,"abstract":"An image preprocessing methodology based on Fourier analysis together with the Laguerre-Gauss Spatial Filter is proposed. This is an alternative to obtain features from aerial images that reduces the feature space significantly, preserving enough information for classification tasks. Experiments on a challenging data set of aerial images show that it is possible to learn a robust classifier from this transformed and smaller feature space using simple models, with similar performance to the complete feature space and more complex models.","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116737812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Breaking the State of the Art in dialects of Spanish Sentiment Analysis","authors":"Daniel Palomino","doi":"10.52591/lxai202007134","DOIUrl":"https://doi.org/10.52591/lxai202007134","url":null,"abstract":"Recent advances in neural language models such as ULMFit and BERT have shown impressive results on several Natural Language Processing (NLP) Tasks. However, this same performance for the Spanish language has not been equaled until now. In this work, we present a combined approach based on training a BERT language model within a modified ULMFit pipeline that allows us to obtain state-of-the-art results Spanish sentiment analysis for several dialects. In order to reinforce our approach we have performed our tests on relevant challenges (Martinez-Camara et al., 2017; 2018).","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130713708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Darwin Danilo Saire Pilco, Adín Ramírez Rivera, S. Tabbone
{"title":"Semantic Segmentation Through Graph Neural Network Blocks","authors":"Darwin Danilo Saire Pilco, Adín Ramírez Rivera, S. Tabbone","doi":"10.52591/lxai202007135","DOIUrl":"https://doi.org/10.52591/lxai202007135","url":null,"abstract":"Semantic segmentation task aims to create a dense classification by labeling pixel-wise each object present in images. Convolutional Neural Network (CNN) approaches have been proved useful by exhibiting the best results in this task. However, some challenges remain, such as the lowresolution of feature maps and the loss of spatial precision, both produced in the CNNs by limited local neighborhoods, i.e., filters with small size. In this work, we propose an encoder-decoder architecture with skip connections based on Graph Neural Network (GNN) (hereafter called GNNblock). This GNN-block proved to have a greater receptive field by having a global vision of objects and their relationships, thus providing additional global information to the model. Finally, we present preliminary results on Cityscape database, achieving close performance with state-of the-art.","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128441945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Deep Clustering Self-Organizing Maps with Relevance Learning","authors":"Pedro H. M. Braga, Heitor Rapela, H. Bassani","doi":"10.52591/lxai202007138","DOIUrl":"https://doi.org/10.52591/lxai202007138","url":null,"abstract":"Nowadays, with the advance of technology, there is an increasing amount of unstructured data being generated every day. However, it is a painful job to label and organize it. Labeling is an expensive, time-consuming, and difficult task. It is usually done manually, which collaborates with the incorporation of noise and errors to the data. Hence, it is of great importance to developing intelligent models that can benefit from both labeled and unlabeled data. Currently, works on unsupervised and semi-supervised learning are still being overshadowed by the successes of purely supervised learning. However, it is expected that they become far more important in the longer term. This article presents a semi-supervised model, called Batch SemiSupervised Self-Organizing Map (Batch SS-SOM), which is an extension of a SOM incorporating some advances that came with the rise of Deep Learning, such as batch training. The results show that Batch SS-SOM is a good option for semisupervised classification and clustering. It performs well in terms of accuracy and clustering error, even with a small number of labeled samples, as well as when presented to unsupervised data, and shows competitive results in transfer learning scenarios in traditional image classification benchmark datasets.","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124220425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Smart Sensor Calibration with Auto-Rotating Perceptrons","authors":"Daniel Saromo, Leonardo Bravo, E. Villota","doi":"10.52591/lxai202007133","DOIUrl":"https://doi.org/10.52591/lxai202007133","url":null,"abstract":"Sensor calibration is vital to have valid measurements of physical activities. Herein we deal with adjusting the signal from a wearable force sensor against a reference scale. By using a few samples and data augmentation, we trained a neural-based regression model to correct the wearable output. For this task, we tested the novel Auto-Rotating Perceptrons (ARP). We found that a neural ARP model with sigmoid activations can outperform an identical neural network based on classic perceptrons with sigmoid and even ReLU activation.","PeriodicalId":402898,"journal":{"name":"LatinX in AI at International Conference on Machine Learning 2020","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122081411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}