Intelligent Models to Identification and Treatment of Outliers in Electrical Load Data

Salgado, Ricardo Menezes; Machado, Tadeu Carvalho; Ohishi, Takaaki

doi:10.1109/tla.2016.7786306

Cited by 6 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Untuk melihat teknik dan pendekatan yang digunakan dalam masing-masing penelitian, dibuatlah klasterisasi berdasarkan bidang yang dapat dilihat pada tabel 4 dibawah. [33] Outlier Load modelling [38] Prediction Local search algorithm [25] Economy Classification Genetic algorithms [27] Prediction Machine Learning [42] Education Association Relationships in learning systems [20] Health Classification Machine learning; Neural Network [39] Medical Outlier Cluster analysis [31] Outlier Data quality assessment [39] Regression Logistic regression; Machine learning; Neural network; Support vector machine [26] Regression Multi-criteria decision analysis; Spatial analysis [45] Social Clustering Hierarchical clustering, K-Medoids, fuzzy clustering, and Self-Organising Maps (SOM) [31] Teknik yang paling banyak digunakan pada penelitian kualitatif adalah clustering dan outlier dapat dilihat pada gambar 4 dibawah. Clustering sendiri merupakan salah satu metode data mining yang bersifat tanpa arahan (unsupervised), maksudnya metode ini diterapkan tanpa adanya latihan (training) dan tanpa ada guru (teacher) serta tidak memerlukan target output.…”

Section: Gambar 3 Sebaran Bidang Penggunaan Data Kualitatifunclassified

Analisis Data Artikel Sistem Pakar Menggunakan Metode Systematic Review

Sastypratiwi

Nyoto

2020

JEPIN

View full text Add to dashboard Cite

Sistem pakar memiliki berbagai keunggulan dibandingkan kepakaran manusia karena sistem pakar terjangkau, permanen, konsisten, proses yang cepat, dan dapat digandakan. Hal ini menyebabkan sistem pakar berkembang diberbagai bidang. Perlu adanya kajian tentang data implementasi maupun metode yang digunakan dalam sistem pakar. Penelitian ini bertujuan untuk menganalisis data pada artikel sistem pakar dengan mengguankan metode systematic review serta menentukan string yang cocok dalam pengumpulan data. Langkah penelitian menggunakan flowchart PRISMA dengan penentuan string yang sesuai berdasarkan lingkup sistem pakar dan teknik dalam data mining. Pencarian dilakukan pada database online. Metode systematic review ini, dapat menjadi salah satu alternatif metode dalam penulisan karya ilmiah berdasarkan artikel sebelumnya. Dimana hasil dalam artikel sistem pakar sebagai studi dalam penelitian bahwa bidang yang masih unggul adalah computer science serta prediction adalah teknik yang paling banyak digunakan.

show abstract

Section: Gambar 3 Sebaran Bidang Penggunaan Data Kualitatifunclassified

Analisis Data Artikel Sistem Pakar Menggunakan Metode Systematic Review

Sastypratiwi

Nyoto

2020

JEPIN

View full text Add to dashboard Cite

show abstract

“…Various research studies have been conducted on a prediction system for power-related data [77][78][79][80][81][82]. In recent years, active research efforts have been made to investigate pattern mining such as power demand and patterns and analyze outlier for the identification of defective data from the collected data [83,84]. This study compares and evaluates the model proposed in the three research studies below to analyze the transmission line tower sensor data of power-related data [85][86][87].…”

Section: Electric Power Prediction Systemmentioning

confidence: 99%

A Novel on Transmission Line Tower Big Data Analysis Model Using Altered K-means and ADQL

Jung

2019

Sustainability

View full text Add to dashboard Cite

This study sought to propose a big data analysis and prediction model for transmission line tower outliers to assess when something is wrong with transmission line tower big data based on deep reinforcement learning. The model enables choosing automatic cluster K values based on non-labeled sensor big data. It also allows measuring the distance of action between data inside a cluster with the Q-value representing network output in the altered transmission line tower big data clustering algorithm containing transmission line tower outliers and old Deep Q Network. Specifically, this study performed principal component analysis to categorize transmission line tower data and proposed an automatic initial central point approach through standard normal distribution. It also proposed the A-Deep Q-Learning algorithm altered from the deep Q-Learning algorithm to explore policies based on the experiences of clustered data learning. It can be used to perform transmission line tower outlier data learning based on the distance of data within a cluster. The performance evaluation results show that the proposed model recorded an approximately 2.29%~4.19% higher prediction rate and around 0.8% ~ 4.3% higher accuracy rate compared to the old transmission line tower big data analysis model.

show abstract

“…Assim, tornam-se indispensáveis modelos de previsão para a produção da energia eólica para auxiliar o sistema elétrico brasileiro, principalmente, no planejamento e na programação energética. Quanto a estes modelos de predição, podemos separá-los, segundo Lei (2009), Foley (2012) e Daraeepour e Echeverri (2014), em 4 metodologias: modelos Research, Society and Development, v. 9, n. 12, e38291211251, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i12.11251 meteorológicos, modelos estatísticos tradicionais (Ruppert, 2011;Eldali, 2016) modelos inteligentes (Liu, 2010;Daraeepour;Echeverri, 2014;Yang, 2015) e modelos híbridos (Chang, 2016;Salgado, 2006;Salgado;Machado, 2016;Siqueira;Salgado, 2011).…”

Section: Introductionunclassified

Previsão de Energia Eólica: Modelo de Ensemble Baseado em Modelos Estatísticos e de Aprendizado de Máquina

Gebin

Salgado

Nogueira

2020

RSD

View full text Add to dashboard Cite

O setor de energia é um dos pilares da sociedade moderna, tendo em vista que pode impactar o desenvolvimento de um país em diversos segmentos – econômico, social, ambiental etc. Quanto ao impacto ambiental, sabe-se que sua produção pode ocasionar resíduos e gerar malefícios para o meio ambiente, ocasionando aumento do efeito estufa e contribuindo para o aquecimento global. Neste sentido, ganha destaque o setor de energia eólica, baseado em uma produção derivada do vento e considerado “limpo”. No entanto, existe certa incerteza em relação a sua produção, por não ser possível produzir ou estocar o vento. Por conta disso, tornam-se necessários modelos de predição para que haja segurança na utilização deste tipo de produção. As predições não são tarefas triviais, visto que existem diversas variáveis que impactam sua obtenção, bem como sua estabilidade ou (instabilidade). Diversos modelos já foram utilizados com este intuito, como os tradicionais (ARIMA), os inteligentes (XGBoost e Random Forest) e os modelos ensemble (junção de modelos), que vêm ganhando muito destaque. O objetivo do trabalho é desenvolver uma estratégia para prever a produção de energia eólica em base horária, em duas estações distintas e com diferentes modelos. Após a realização dos experimentos, pode-se perceber que os modelos individuais (com destaque para o XGBoost) apresentaram bons resultados; no entanto, deve ser notado que, na maioria dos casos de picos/outliers, estes modelos sofrem de superestimação. O modelo ensemble, entretanto, conseguiu suprir tal deficiência em relação aos modelos individuais, corrigindo os casos de superestimação.

show abstract

Intelligent Models to Identification and Treatment of Outliers in Electrical Load Data

Cited by 6 publications

References 7 publications

Analisis Data Artikel Sistem Pakar Menggunakan Metode Systematic Review

Analisis Data Artikel Sistem Pakar Menggunakan Metode Systematic Review

A Novel on Transmission Line Tower Big Data Analysis Model Using Altered K-means and ADQL

Previsão de Energia Eólica: Modelo de Ensemble Baseado em Modelos Estatísticos e de Aprendizado de Máquina

Contact Info

Product

Resources

About