Endah Kristiani

Dr. Endah Kristiani

orcid

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 July 2019 in Future Generation Computer Systems
Verified
Journal Article
Closed access
Chao-Tung YangChao-Tung Yang
Chao-Tung Yang
SCShuo-Tsung Chen
Shuo-Tsung Chen
Walter DenWalter Den
Walter Den
YWYun-Ting Wang
Yun-Ting Wang
EKEndah Kristiani
Endah Kristiani

11 February 2022 in Sustainability
Verified
Journal Article
Open access
PDF file
EKEndah Kristiani
Endah Kristiani
HLHao Lin
Hao Lin
JLJwu-Rong Lin
Jwu-Rong Lin
Yen-Hsun ChuangYen-Hsun Chuang
Yen-Hsun Chuang
CHChin-Yin Huang
Chin-Yin Huang
Chao-Tung YangChao-Tung Yang
Chao-Tung Yang
This paper implements deep learning methods of recurrent neural networks and short-term memory models. Two kinds of time-series data were used: air pollutant factors, such as O3, SO2, and CO2 from 2017 to 2019, and meteorological factors such as temperature, humidity, wind direction, and wind speed. A trained model was used to predict air pollution within an eight-hour period. Correlation analysis was applied using Pearson and Spearman correlation coefficients. The KNN method was used to fill in the missing values to improve the generated model’s accuracy. The average absolute error percentage value was used in the experiments to evaluate the model’s performance. LSTM had the lowest RMSE value at 1.9 than the other models from the experiments. CNN had a significant RMSE value at 3.5, followed by Bi-LSTM at 2.5 and Bi-GRU at 2.7. In comparison, the RNN was slightly higher than LSTM at a 2.4 value.

YTYu-Tse Tsan
Yu-Tse Tsan
Der-Yuan ChenDer-Yuan Chen
Der-Yuan Chen
PLPo-Yu Liu
Po-Yu Liu
EKEndah Kristiani
Endah Kristiani
Kieu Lan Phuong NguyenKieu Lan Phuong Nguyen
Kieu Lan Phuong Nguyen
Chao-Tung YangChao-Tung Yang
Chao-Tung Yang
This paper proposed the forecasting model of Influenza-like Illness (ILI) and respiratory disease. The dataset was extracted from the Taiwan Environmental Protection Administration (EPA) for air pollutants data and the Centers for Disease Control (CDC) for disease cases from 2009 to 2018. First, this paper applied the ARIMA method, which trained based on the weekly number of disease cases in time series. Second, we implemented the Long short-term memory (LSTM) method, which trained based on the correlation between the weekly number of diseases and air pollutants. The models were also trained and evaluated based on five and ten years of historical data. Autoregressive integrated moving average (ARIMA) has an excellent model in the five-year dataset of ILI at 2564.9 compared to ten years at 8173.6 of RMSE value. This accuracy is similar to the Respiratory dataset, which gets 15,656.7 in the five-year dataset and 22,680.4 of RMSE value in the ten-year dataset. On the contrary, LSTM has better accuracy in the ten-year dataset than the five-year dataset. For example, on average of RMSE in the ILI dataset, LSTM has 720.2 RMSE value in five years and 517.0 in ten years dataset. Also, in the Respiratory disease dataset, LSTM gets 4768.6 of five years of data and 3254.3 of the ten-year dataset. These experiments revealed that the LSTM model generally outperforms ARIMA by three to seven times higher model performance.

1 October 2021 in Journal of Hazardous Materials
Verified
Journal Article
Closed access
CYChao-Tung Yang
Chao-Tung Yang
HCHo-Wen Chen
Ho-Wen Chen
ECEn-Jui Chang
En-Jui Chang
EKEndah Kristiani
Endah Kristiani
KNKieu Lan Phuong Nguyen
Kieu Lan Phuong Nguyen
JCJo-Shu Chang
Jo-Shu Chang

21 January 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
CLChun-Ming Lai
Chun-Ming Lai
MCMei-Hua Chen
Mei-Hua Chen
EKEndah Kristiani
Endah Kristiani
Vinod Kumar VermaVinod Kumar Verma
Vinod Kumar Verma
Chao-Tung YangChao-Tung Yang
Chao-Tung Yang
Due to the openness and easy accessibility of online social media (OSM), anyone can easily contribute a simple paragraph of text to express their opinion on an article that they have seen. Without access control mechanisms, it has been reported that there are many suspicious messages and accounts spreading across multiple platforms. Accordingly, identifying and labeling fake news is a demanding problem due to the massive amount of heterogeneous content. In essence, the functions of machine learning (ML) and natural language processing (NLP) are to enhance, speed up, and automate the analytical process. Therefore, this unstructured text can be transformed into meaningful data and insights. In this paper, the combination of ML and NLP are implemented to classify fake news based on an open, large and labeled corpus on Twitter. In this case, we compare several state-of-the-art ML and neural network models based on content-only features. To enhance classification performance, before the training process, the term frequency-inverse document frequency (TF-IDF) features were applied in ML training, while word embedding was utilized in neural network training. By implementing ML and NLP methods, all the traditional models have greater than 85% accuracy. All the neural network models have greater than 90% accuracy. From the experiments, we found that the neural network models outperform the traditional ML models by, on average, approximately 6% precision, with all neural network models reaching up to 90% accuracy.

4 February 2020 in IEEE Access
Verified
Journal Article
Open access
EKEndah Kristiani
Endah Kristiani
Chao-Tung YangChao-Tung Yang
Chao-Tung Yang
CHChin-Yin Huang
Chin-Yin Huang
Optimization strategies in deep learning models require different techniques for different use cases. Besides, various phases of the model deployment life-cycle specify possible and particular optimization strategies. In this paper, an optimized deep learning model on the edge computing environment is proposed for image classification cases. For preparing the dataset, the image preprocessing and data augmentation methods are utilized to prepare the data for the training process. To accelerate the deep learning training process, this system implemented CPU optimization and hyperparameter tuning. Tensorflow is applied as a framework for the training model. InceptionV3, VGG16, and MobileNet are applied as topology implemented in the deep learning training comparison. In this case, InceptionV3 was used for modeling the deep learning applications on edge. To optimize the trained model, a Model Optimizer is used on the edge device. It can be seen in the experiments, MobileNet was the least accurate model (85%) and the longest time to load the model (71s). VGG16 was the most reliable (91%) and the shortest time to load the model (50s). InceptionV3 has median accuracy (87%) and the average time to load the model (52s).