Search

In this paper, we present a framework for performing automatic analysis of Land Use Zones based on Location-Based Social Networks (LBSNs). We model city areas using a hierarchical structure of POIs extracted from foursquare. We encode such structures in kernel machines, e.g., Support Vector Machines, using a new Tree Kernel, i.e., the Hierarchical POI Kernel (HPK), which can take the importance of the individual POIs into account during the substructure matching. This way, HPK projects structures in the space of all their possible substructures such that each dimension corresponds to a semantic structural feature, weighted according to the discriminative power of POIs. We generated four different datasets for the following cities: Barcelona, Lisbon, Amsterdam and Milan, where we trained and tested our models. The results show that our approach largely outperforms previous work and standard baseline built on simple features, such as counts of different POIs. Finally, we apply a mining algorithm to extract the most relevant features (tree fragments) from the implicit TK space according to the weights the kernel machine assigned to them. Our approach can produce an explicit set of representative features that can be used to classify and characterize urban areas. [ABSTRACT FROM AUTHOR]

This paper introduces a methodology based on Python libraries and machine learning k-Nearest Neighbors (KNN) algorithms to create an interactive 3D HTML model (3D_Vertical_Sections_Faults_LRD.html) that combines 2D grain-size KNN-prediction vertical maps (vertical sections) from which syn-sedimentary faults and other features in sedimentary porous media can be delineated. The model can be visualized and handled with conventional web browsers. The grain-size physical parameter is measurable, constant over instrumental time, handleable mathematically, and its range can be associated to lithological classes. Grain-size input data comes from a public database of 433 boreholes in the Llobregat River Delta (LRD) in NE Spain. Four lithological classes were defined: Pre-Quaternary basement, and Quaternary gravel, sand, and clay–silt. Using a new KNN-prediction algorithm, seven NW–SE (transversal) and three SW–NE (longitudinal) vertical sections were created following the orientation of faults identified in surface and detected in reflection seismic geophysical surveys. For exploratory K values in the 1–75 range were used. K around 25 provides the general and smoothy shape of the basement top surface, whereas K = 1 is a optimal value to represent the heterogeneity of the LRD at short distance. Using a new KNN-prediction confidence algorithm inspired in the Similarity Ratio algorithm for machine-learning KNN, the vertical sections overall confidence was evaluated as satisfactory. A general decreasing confidence trend according to the decreasing data density with depth and from inland to seaward was found. The vertical sections created with K = 1 show horizontal interruptions (displacements or vertical steps) in the basement continuity and in the Quaternary coarse bodies (gravel and sand) attributable to the action of Quaternary active faults. These faults have been linked or correlated with well-known active faults in the area related in much cases with the Valencia Trough opening. Moreover, several faults detected in surface and other identified in this paper by the first time have been revealed as fault zones made of fault branches with different steps in an echelon-like arrangement. Faulting seems to be more evident in the Pleistocene Lower Detrital Complex and much less active or inactive in the Holocene Upper Detrital Complex. Syn-tectonic gravel channels faulty controlled, progradation of gravel lobes, and lateral migration of channel bars were also observed. At its current development stage, this methodology could also be applied to other geological environments, making the due minor modifications of the code, and is especially suitable to reduce the high (usually unmeasurable) uncertainty associated to the qualitative geological data used in more complex numerical tools aimed at modelling a lot of geological resources (groundwater, minerals, geothermal, petroleum) or different Earth phenomena. • Llobregat River Delta (Barcelona, NE Spain). • Python Libraries. • K-Nearest Neighbors algorithm. • D HTML models. • Syn-sedimentary faults and gravel channels migration. [ABSTRACT FROM AUTHOR]

This paper studies how a group of delivery workers in Barcelona were able to organize a successful traditional and a social media strategy in order to claim for their rights as waged workers. They created a union, RidersxDerechos, and they also decided to create a worker's cooperative, Mensakas, with their own application and algorithm. We will study how they were able to re-appropriate technology and to use digital communities to spread alternative discourses. We have used different methodologies: traditional content analysis in Media, debate analysis in Social Media, qualitative ethnography. We noticed that RidersxDerechos access to media was very successful (300 piece of news analyzed) thanks to strikes and court trials, facilitating a change of perspective in the treatment of platform economy in Media. Media were following up the digital entrepreneurship rhetoric until then. Along with the traditional media strategy they developed a diversified communicative pathway in social media (Twitter, Instagram, Facebook, and Goteo) that helped them to establish alliances with riders from other cities and countries. We focused on more than 25.000 tweets. Finally, they proposed a new way to use technology by creating their own app and algorithms for their working cooperative, Mensakas. Crowdfunding was also used to fund it and to spread an alternative working storytelling from Silicon Valley's. [ABSTRACT FROM AUTHOR]

Abstract: This paper presents a signal analysis methodology to validate (detect) and reconstruct the missing and false data of a large set of flow meters in the telecontrol system of a water distribution network. The proposed methodology is based on two time-scale forecasting models: a daily model based on a ARIMA time series, while the 10-min model is based on distributing the daily flow using a 10-min demand pattern. The demand patterns have been determined using two methods: correlation analysis and an unsupervised fuzzy logic classification, named LAMDA algorithm. Finally, the proposed methodology has been applied to the Barcelona water distribution network, providing very good results. [Copyright &y& Elsevier]

To automatically detect dynamic EEG signals to reduce the time cost of epilepsy diagnosis. In the signal recognition of electroencephalogram (EEG) of epilepsy, traditional machine learning and statistical methods require manual feature labeling engineering in order to show excellent results on a single data set. And the artificially selected features may carry a bias, and cannot guarantee the validity and expansibility in real-world data. In practical applications, deep learning methods can release people from feature engineering to a certain extent. As long as the focus is on the expansion of data quality and quantity, the algorithm model can learn automatically to get better improvements. In addition, the deep learning method can also extract many features that are difficult for humans to perceive, thereby making the algorithm more robust. Based on the design idea of ResNeXt deep neural network, this paper designs a Time-ResNeXt network structure suitable for time series EEG epilepsy detection to identify EEG signals. The accuracy rate of Time-ResNeXt in the detection of EEG epilepsy can reach 91.50%. The Time-ResNeXt network structure produces extremely advanced performance on the benchmark dataset (Berne-Barcelona dataset) and has great potential for improving clinical practice. [ABSTRACT FROM AUTHOR]