Transfer Learning for Node Regression Applied to Spreading Prediction

Understanding how information propagates in real-life complex networks yields a better understanding of dynamic processes such as misinformation or epidemic spreading. The recently introduced branch of machine learning methods for learning node representations offers many novel applications, one of them being the task of spreading prediction addressed in this paper. We explore the utility of the state-of-the-art node representation learners when used to assess the effects of spreading from a given node, estimated via extensive simulations. Further, as many real-life networks are topologically similar, we systematically investigate whether the learned models generalize to previously unseen networks, showing that in some cases very good model transfer can be obtained. This paper is one of the first to explore transferability of the learned representations for the task of node regression; we show there exist pairs of networks with similar structure between which the trained models can be transferred (zero-shot) and demonstrate their competitive performance. To our knowledge, this is one of the first attempts to evaluate the utility of zero-shot transfer for the task of node regression.

Exploring model transfer strategies for sentiment analysis in Twitter

As mídias sociais se tornaram um ambiente popular para comunicação. Por isso, analisar o sentimento que o usuário expressa em suas postagens nas redes sociais é um importante campo de pesquisa. No entanto, detectar a polaridade em tais conteúdos é um desafio, em parte porque a quantidade de dados rotulados para treinar classificadores é escassa em muitas situações. Este artigo explora estratégias para reusar um modelo aprendido a partir de conjunto de dados fonte para classificar instâncias em um conjunto de dados de destino. Os experimentos são conduzidos com 22 conjuntos de dados de análise de sentimento em tweets e abordagens baseadas em métricas de similaridade. Os resultados apontam que o tamanho do conjunto de treinamento fonte desempenha um papel essencial no desempenho dos classificadores quando usados para inferir a classe das instâncias alvo.

Penerapan transfer learning pada convolutional neural networks dalam deteksi covid-19.

Pandemi Covid-19 menjadi masalah serius di Dunia termasuk Indonesia sampai saat ini, virus yang muncul pada akhir tahun 2019 ini masih menjadi masalah serius. Jumlah kasus orang yang terinfeksi terus meningkat dan mencapai angka lebih dari dua ratus juta kasus di seluruh dunia. Untuk melakukan tes cepat ini tidak langsung berjalan dengan lancar tetapi mengalami banyak kendala yang dialami oleh tim Medis, salah satunya keterbatasan kit tes Covid-19, sehingga ilmuwan mengambil langkah diagnosis lainnya. Dalam bidang informatika ilmuwan banyak menggunakan beberapa diagnosis salah satunya gambar X-ray pada paru-paru. Gambar CXR pada saat ini sering digunakan untuk proses deteksi menggunakan algoritma CNN. Penelitian ini menggunakan metode transfer learning yang akan diuji dalam dataset skala besar dan kecil. Hasil terbaik dari semua model yang dicoba yaitu MobileNet dengan hasil akurasi 98.11% yang diuji pada dataset skala besar dan paling rendah didapat oleh ResNet50 yang diuji pada dataset skala kecil dengan akurasi 41.94%. Dataset dalam skala besar juga menjunjukkan peningkatan akurasi pada semua model transfer learning yang diuji.

Penerapan transfer learning pada convolutional neural networks dalam deteksi covid-19.

Pandemi Covid-19 menjadi masalah serius di Dunia termasuk Indonesia sampai saat ini, virus yang muncul pada akhir tahun 2019 ini masih menjadi masalah serius. Jumlah kasus orang yang terinfeksi terus meningkat dan mencapai angka lebih dari dua ratus juta kasus di seluruh dunia. Untuk melakukan tes cepat ini tidak langsung berjalan dengan lancar tetapi mengalami banyak kendala yang dialami oleh tim Medis, salah satunya keterbatasan kit tes Covid-19, sehingga ilmuwan mengambil langkah diagnosis lainnya. Dalam bidang informatika ilmuwan banyak menggunakan beberapa diagnosis salah satunya gambar X-ray pada paru-paru. Gambar CXR pada saat ini sering digunakan untuk proses deteksi menggunakan algoritma CNN. Penelitian ini menggunakan metode transfer learning yang akan diuji dalam dataset skala besar dan kecil. Hasil terbaik dari semua model yang dicoba yaitu MobileNet dengan hasil akurasi 98.11% yang diuji pada dataset skala besar dan paling rendah didapat oleh ResNet50 yang diuji pada dataset skala kecil dengan akurasi 41.94%. Dataset dalam skala besar juga menjunjukkan peningkatan akurasi pada semua model transfer learning yang diuji.

Unsupervised cross-lingual model transfer for named entity recognition with contextualized word representations

Named entity recognition (NER) is one fundamental task in the natural language processing (NLP) community. Supervised neural network models based on contextualized word representations can achieve highly-competitive performance, which requires a large-scale manually-annotated corpus for training. While for the resource-scarce languages, the construction of such as corpus is always expensive and time-consuming. Thus, unsupervised cross-lingual transfer is one good solution to address the problem. In this work, we investigate the unsupervised cross-lingual NER with model transfer based on contextualized word representations, which greatly advances the cross-lingual NER performance. We study several model transfer settings of the unsupervised cross-lingual NER, including (1) different types of the pretrained transformer-based language models as input, (2) the exploration strategies of the multilingual contextualized word representations, and (3) multi-source adaption. In particular, we propose an adapter-based word representation method combining with parameter generation network (PGN) better to capture the relationship between the source and target languages. We conduct experiments on a benchmark ConLL dataset involving four languages to simulate the cross-lingual setting. Results show that we can obtain highly-competitive performance by cross-lingual model transfer. In particular, our proposed adapter-based PGN model can lead to significant improvements for cross-lingual NER.

Relay Identification Using Shifting Method for PID Controller Tuning

The aim of this study was to present a relay shifting method for relay feedback identification of dynamical systems suitable for PID controller tuning. The proposed technique uses a biased relay to determine frequency response points from a single experiment without any assumptions about a model transfer function. The method is applicable for open-loop stable, unstable, and integration processes, even with a delay, and regardless of whether they are oscillating or non-oscillating. The core of this technique was formed by the so-called relay shifting filter. In this study, the method was applied to a parameter estimation of a second-order time-delayed (SOTD) model that can describe, with acceptable accuracy, the dynamics of most processes (even with a transport delay) near the operating point. Simultaneously, a parameter setting for the PID controller was derived based on the model parameters. The applicability of the proposed method was demonstrated on various simulated processes and tested on real laboratory apparatuses.

FAST.Farm load validation for single wake situations at alpha ventus

The main objective of the presented work is the validation of the simulation tool FAST.Farm for the calculation of power and structural loads in single wake situations; the basis for the validation is the measurement database of the operating offshore wind farm alpha ventus. The approach is described in detail and covers the calibration of the aeroelastic turbine model, transfer of environmental conditions to simulations, and comparison between simulations and adequately filtered measurements. It is shown that FAST.Farm accurately predicts power and structural load distributions over wind direction with discrepancies of less than 10 % for most of the cases compared to the measurements. Additionally, the frequency response of the structure is investigated, and it is calculated by FAST.Farm in good agreement with the measurements. In general, the calculation of fatigue loads is improved with a wake-added turbulence model added to FAST.Farm in the course of this study.