1,720,968 research outputs found
Multi-label classification with imbalanced classes by fuzzy deep neural networks
No full textMulti-label classification is an advantageous technique for managing uncertainty in classification problems where each data instance is associated with several labels simultaneously. Such situations are frequent in real-world scenarios, where decisions rely on imprecise or noisy data and adaptable classification methods are preferred. However, the problem of class imbalance represents a common characteristic of several multi-label datasets, in which the distribution of samples and their corresponding labels is non-uniform across the data space. In this paper, we propose a multi-label classification approach utilizing fuzzy logic in order to deal with the class imbalance problem. To eliminate the need for an expert to determine the logical rules of inference, deep neural networks are adopted, which have proven to be exceptionally effective for such problems. By combining both fuzzy inference systems and deep neural networks, the strengths and weaknesses of each approach can be mitigated. As a further development, a symbolic representation of time series is put in place to reduce data dimensionality and speed up the training procedure. This allows for more flexibility in model application, in particular with respect to time constraints arising from the causality of observed time series. Tests carried out on a multi-label classification dataset related to the current and voltage profiles of several household appliances show that the proposed model outperforms four baseline models for time series classification
An adaptive embedding procedure for time series forecasting with deep neural networks
Full text linkNowadays, solving time series prediction problems is an open and challenging task. Many solutions are based on the implementation of deep neural architectures, which are able to analyze the structure of the time series and to carry out the prediction. In this work, we present a novel deep learning scheme based on an adaptive embedding mechanism. The latter is exploited to extract a compressed representation of the input time series that is used for the subsequent forecasting. The proposed model is based on a two-layer bidirectional Long Short-Term Memory network, where the first layer performs the adaptive embedding and the second layer acts as a predictor. The performances of the proposed forecasting scheme are compared with several models in two different scenarios, considering both well-known time series and real-life application cases. The experimental results show the accuracy and the flexibility of the proposed approach, which can be used as a prediction tool for any actual application
Multivariate time series analysis for electrical power theft detection in the distribution grid
No full textClassification of time series is a fundamental problem in energy distribution, especially to extract information about events that occurred during the observation period. In this paper, we propose a solution to the problem of identifying energy thefts by introducing a classification method based on convolutional neural networks. The input structure to the model is based on real data that have been certified by external authorities and regards thefts operated by the final user with physical intervention. The training of the neural network is done by means of yearly time series of monthly data, which pertain to different physical quantities relevant to the user profile. The proposed method has been experimentally tested and verified against acceptable test results in different conditions, even giving an indication on where in the sequence the theft has occurred
ADMM consensus for deep LSTM networks
No full textIn modern real-world applications, the need of using a decentralized data processing approach has progressively increased, facing complexity and handling issues. Pervasive data and ubiquitous computational capacity have enabled the proficient use of distributed implementation of machine learning algorithms, especially for forecasting problems. We provide in this paper a new, fully distributed prediction approach based on the Long Short-Term Memory deep neural network. When placed in a network of interconnected agents, the single predictors are able to improve the prediction accuracy by means of the Alternating Direction Method of Multipliers consensus procedure on some network parameters. Experimental tests on real-world time series prove the efficacy of the proposed approach, which regulates the information exchange in the network through high-level structures in the considered models
A neural network symbolic approach to structural health monitoring in aerospace applications
No full textDeep Learning models, and specifically Recurrent Neural Networks, have been successfully applied to time series classification in many applications, including Structural Health Monitoring. A relatively new field of research is the implementation of Deep Learning techniques for damage identification via measured time series in space systems, which proves chal-lenging in case of Structural Health Monitoring of large flexible structures. In this work, we propose a novel approach exploiting a symbolic time series representation as an additional data pre-processing step for data dimensionality reduction. The strategy is applied to a real world-scenario of a spacecraft hosting large solar panels equipped with distributed accelerometers at structural level, and compared against a previous benchmark case developed by the authors. Obtained results prove that the strategy has the potential to further improve classification quality towards an ideal 100% accuracy envisioned for space systems
A blockwise embedding for multi-day-ahead prediction of energy time series by randomized deep neural networks
No full textNowadays, deep learning is gaining attraction as one of the most successful paradigm for a plethora of machine learning applications. While its benefits are undoubted, the high computational burden associated with its training algorithms and cross-validation procedures is stimulating new lines of research. To this end, randomized deep neural networks are one of the best alternatives in terms of efficiency-to-accuracy balance. In this paper, we present a deep neural architecture that uses randomization of some parameters in a complex structure whose novelty is twofold: it embeds past samples of the time series by using daily blocks in the input frame of a convolutional layer; it predicts a day on the whole by solving a suitable regression problem. The proposed randomized approach is compared with state-of-the-art prediction algorithms on the challenging context related to energy time series, where the day-ahead prediction is usual, obtaining comparable or even better results in terms of forecasting accuracy and training time
Nonexclusive classification of household appliances by fuzzy deep neural networks
No full textFuzzy classification is a very useful tool for managing the uncertainty in a classification problem with non-mutually exclusive classes, whose values can fall into overlapping ranges. This situation is very common in real-life problems, where decisions are often made on the basis of inaccurate or noisy information and a flexible classification is preferred. In this paper, we propose a nonexclusive classification approach based on fuzzy logic to classify household appliances characterized by the time series associated with their power consumption. This issue is crucial for purposes related to user profiling, demand side management and cost optimization in the context of smart grids and green energy communities. To overcome the dependence on an expert for determining the logical rules of inference, we rely on the use of deep neural networks, as they have proved to be an extremely powerful tool in this kind of problems. The advantages and disadvantages present in fuzzy inference systems and deep neural networks almost completely disappear when both models are combined. In this regard, the paper proposes a randomization-based fuzzy deep neural network for the nonexclusive classification of household appliances. Randomization in deep neural networks allows a significant reduction in training times while often maintaining a high level of precision. This enables the adopted model with respect to time constraints causality of the observed time series. The performances obtained from the proposed model compare favorably with those obtained using two benchmark models for time series classification based on the well-known Long Short-Term Memory network
2-d convolutional deep neural network for the multivariate prediction of photovoltaic time series
No full textHere, we propose a new deep learning scheme to solve the energy time series prediction problem. The model implementation is based on the use of Long Short-Term Memory networks and Convolutional Neural Networks. These techniques are combined in such a fashion that interdependencies among several different time series can be exploited and used for forecasting purposes by filtering and joining their samples. The resulting learning scheme can be summarized as a superposition of network layers, resulting in a stacked deep neural architecture. We proved the accuracy and robustness of the proposed approach by testing it on real-world energy problems
Multivariate Prediction of Energy Time Series by Autoencoded LSTM Networks
No full textIn this paper, a novel approach for the multivariate prediction of energy time series is presented. It is based on the Long Short-Term Memory deep neural network. The latter is made up of two stacked recurrent layers and it is used in two different training configurations. First, an encoder-decoder structure is implemented in order to extract meaningful representative features from the time series. Then, this embedded data are used to improve the actual prediction. To prove the goodness of our approach, its performance is compared with two different benchmarks. The numerical results show that the proposed model outperforms the aforementioned benchmarks
A Fast Deep Learning Technique for Wi-Fi-Based Human Activity Recognition
Full text linkDespite recent advances, fast and reliable Human Activity Recognition in confined space is still an open problem related to many real-world applications, especially in health and biomedical monitoring. With the ubiquitous presence of Wi-Fi networks, the activity recognition and classification problems can be solved by leveraging some characteristics of the Channel State Information of the 802.11 standard. Given the well-documented advantages of Deep Learning algorithms in solving complex pattern recognition problems, many solutions in Human Activity Recognition domain are taking advantage of those models. To improve the time and precision of activity classification of time-series data stemming from Channel State Information, we propose herein a fast deep neural model encompassing concepts not only from state-of-the-art recurrent neural networks, but also using convolutional operators with added randomization. Results from real data in an experimental environment show promising results
- …
