Search results for "machine learning."

McRock at SemEval-2022 Task 4: Patronizing and Condescending Language Detection using Multi-Channel CNN, Hybrid LSTM, DistilBERT and XLNet

2022

In this paper we propose four deep learning models for the task of detecting and classifying Patronizing and Condescending Language (PCL) using a corpus of over 13,000 annotated paragraphs in English. The task, hosted at SemEval-2022, consists of two different subtasks. The Subtask 1 is a binary classification problem. Namely, given a paragraph, a system must predict whether or not it contains any form of PCL. The Subtask 2 is a multi-label classification task. Given a paragraph, a system must identify which PCL categories express the condescension. A paragraph might contain one or more categories of PCL. To face with the first subtask we propose a multi-channel Convolutional Neural Network…

A Structural Approach to Infer Recurrent Relations in Data

2014

Extracting knowledge from a great amount of collected data has been a key problem in Artificial Intelligence during the last decades. In this context, the word "knowledge" refers to the non trivial new relations not easily deducible from the observation of the data. Several approaches have been used to accomplish this task, ranging from statistical to structural methods, often heavily dependent on the particular problem of interest. In this work we propose a system for knowledge extraction that exploits the power of an ontology approach. Ontology is used to describe, organise and discover new knowledge. To show the effectiveness of our system in extracting and generalising the knowledge emb…

A machine learning approach for user localization exploiting connectivity data

2016

The growing popularity of Location-Based Services (LBSs) has boosted research on cheaper and more pervasive localization systems, typically relying on such monitoring equipment as Wireless Sensor Networks (WSNs), which allow to re-use the same instrumentation both for monitoring and for localization without requiring lengthy off-line training. This work addresses the localization problem, exploiting knowledge acquired in sample environments, and extensible to areas not considered in advance. Localization is turned into a learning problem, solved by a statistical algorithm. Additionally, parameter tuning is fully automated thanks to its formulation as an optimization problem based only on co…

Resilient hexapod robot

2017

In this paper, we present a method of learning desired behaviour of the specific robotic system and transfer of the existing knowledge in the event of partial system failure. Six-legged robot (hexapod) built on top of the Bioloid platform is used for the method verification. We use genetic algorithms to optimize the hexapod's gait, after which we simulate physical damage caused to the robot. The goal of this method is to optimize the gait in accordance with the actual robot morphology, instead of the assumed one. Also, knowledge that was previously gained will be transferred in order to improve the results. Nonstandard genetic algorithm with the specific mixed population is used for this.

Fast Training of Self Organizing Maps for the Visual Exploration of Molecular Compounds

2007

Visual exploration of scientific data in life science\ud area is a growing research field due to the large amount of\ud available data. The Kohonen’s Self Organizing Map (SOM) is\ud a widely used tool for visualization of multidimensional data.\ud In this paper we present a fast learning algorithm for SOMs\ud that uses a simulated annealing method to adapt the learning\ud parameters. The algorithm has been adopted in a data analysis\ud framework for the generation of similarity maps. Such maps\ud provide an effective tool for the visual exploration of large and\ud multi-dimensional input spaces. The approach has been applied\ud to data generated during the High Throughput Screening\ud of mo…

Structural Knowledge Extraction and Representation in Sensory Data

During the last decades the availability of increasingly cheaper technology for pervasive monitoring has boosted the creation of systems able to automatically comprehend the events occurring in the monitored area, in order to plan a set of actions to bring the environment closer to the user's preferences. These systems must inevitably process a great amount of raw data - sensor measurements - and need to summarize them in a high-level representation to accomplish their tasks. An implicit requirement is the need to learn from experience, in order to be able to capture the hidden structure of the data, in terms of relations between its key components. The availability of large collections of …

Fake News Spreaders Detection: Sometimes Attention Is Not All You Need

2022

Guided by a corpus linguistics approach, in this article we present a comparative evaluation of State-of-the-Art (SotA) models, with a special focus on Transformers, to address the task of Fake News Spreaders (i.e., users that share Fake News) detection. First, we explore the reference multilingual dataset for the considered task, exploiting corpus linguistics techniques, such as chi-square test, keywords and Word Sketch. Second, we perform experiments on several models for Natural Language Processing. Third, we perform a comparative evaluation using the most recent Transformer-based models (RoBERTa, DistilBERT, BERT, XLNet, ELECTRA, Longformer) and other deep and non-deep SotA models (CNN,…

Prioritisation of alternatives with analytical hierarchy process plus response latency and web surveys

2014

This paper introduces a new method that combines the well-known analytical hierarchy process (AHP) with a response latency metric. The response latency is the time taken by respondents to make choices over pairwise comparisons. The analytical calculation of relative importance weights of the alternatives is made by using a response latency model previously validated in several case studies. This combination aims to overcome some drawbacks of the traditional AHP related to the use of a rating scale (the so-called Saaty scale), and it is a natural way to involve response latency in established decisionmaking methods. This new method can be profitably adopted in web surveys where it is easy to…

Discovery of novel trichomonacidals using LDA-driven QSAR models and bond-based bilinear indices as molecular descriptors

2008

Few years ago, the World Health Organization estimated the number of adults with trichomoniasis at 170 million worldwide, more than the combined numbers for gonorrhea, syphilis, and chlamydia. To combat this sexually transmitted disease, Metronidazole (MTZ) has emerged, since 1959, as a powerful drug for the systematic treatment of infected patients. However, increasing resistance to MTZ, adverse effects associated to high-dose MTZ therapies and very expensive conventional technologies related to the development of new trichomonacidals necessitate novel computational methods that shorten the drug discovery pipeline. Therefore, bond-based bilinear indices, new 2-D bond-based TOMOCOMD-CARDD M…

Automatic differentiation of melanoma from dysplastic nevi.

2015

International audience; Malignant melanoma causes the majority of deaths related to skin cancer. Nevertheless, it is the most treatable one, depending on its early diagnosis. The early prognosis is a challenging task for both clinicians and dermatologist, due to the characteristic similarities of melanoma with other skin lesions such as dysplastic nevi. In the past decades, several computerized lesion analysis algorithms have been proposed by the research community for detection of melanoma. These algorithms mostly focus on differentiating melanoma from benign lesions and few have considered the case of melanoma against dysplastic nevi. In this paper, we consider the most challenging task a…