Feature selection based on term frequency deviation rate for text classification

Zhou, Huanfu; Ma, Yiming; Li, Xiang

doi:10.1007/s10489-020-01937-4

Cited by 13 publications

(4 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Macro-F1 score is defined as shown in Eq. [10] ( 40 , 41 ). The Macro-F1 score also lies between 0.0 and 1.0, with the smallest value [0] indicating the worst performance of the classifier, and the highest value [1] indicating the best performance of the classifier.…”

Section: Resultsmentioning

confidence: 99%

A difficulty-aware and task-augmentation method based on meta-learning model for few-shot diabetic retinopathy classification

Liu,

Dong,

et al. 2024

Quant Imaging Med Surg

View full text Add to dashboard Cite

Background Accurate classification techniques are essential for the early diagnosis and treatment of patients with diabetic retinopathy (DR). However, the limited amount of annotated DR data poses a challenge for existing deep-learning models. This article proposes a difficulty-aware and task-augmentation method based on meta-learning (DaTa-ML) model for few-shot DR classification with fundus images. Methods The difficulty-aware (Da) method operates by dynamically modifying the cross-entropy loss function applied to learning tasks. This methodology has the ability to intelligently down-weight simpler tasks, while simultaneously prioritizing more challenging tasks. These adjustments occur automatically and aim to optimize the learning process. Additionally, the task-augmentation (Ta) method is used to enhance the meta-training process by augmenting the number of tasks through image rotation and improving the feature-extraction capability. To implement the expansion of the meta-training tasks, various task instances can be sampled during the meta-training stage. Ultimately, the proposed Ta method was introduced to optimize the initialization parameters and enhance the meta-generalization performance of the model. The DaTa-ML model showed promising results by effectively addressing the challenges associated with few-shot DR classification. Results The Asia Pacific Tele-Ophthalmology Society (APTOS) 2019 blindness detection data set was used to evaluate the DaTa-ML model. The results showed that with only 1% of the training data (5-way, 20-shot) and a single update step (training time reduced by 90%), the DaTa-ML model had an accuracy rate of 89.6% on the test data, which is a 1.7% improvement over the transfer-learning method [i.e., residual neural network (ResNet)50 pre-trained on ImageNet], and a 16.8% improvement over scratch-built models (i.e., ResNet50 without pre-trained weights), despite having fewer trainable parameters (the parameters used by the DaTa-ML model are only 0.47% of the ResNet50 parameters). Conclusions The DaTa-ML model provides a more efficient DR classification solution with little annotated data and has significant advantages over state-of-the-art methods. Thus, it could be used to guide and assist ophthalmologists to determine the severity of DR.

show abstract

Section: Resultsmentioning

confidence: 99%

A difficulty-aware and task-augmentation method based on meta-learning model for few-shot diabetic retinopathy classification

Liu,

Dong,

et al. 2024

Quant Imaging Med Surg

View full text Add to dashboard Cite

show abstract

“…The micro-averaged precision ( miP ) and micro-averaged recall ( miR ) are defined as

Note that for both miP and miR , the denominator is the sum of all the elements (diagonal and off-diagonal) of the confusion matrix, and it is 1. Finally, the micro-averaged F 1 score is defined as the harmonic mean of these quantities:

This definition is commonly used (e.g., [ 6 , 8 – 12 , 14 , 15 ]).…”

Section: Averaged F 1 Scoresmentioning

confidence: 99%

“…This version of macro-averaged F 1 score is less frequently used (e.g., [ 11 , 12 , 16 ]). For our example,

In this example, the micro-averaged F 1 score is higher than the macro-averaged F 1 scores because both within-class precision and recall are much lower for the first class compared to the other two.…”

Section: Averaged F 1 Scoresmentioning

confidence: 99%

Confidence interval for micro-averaged F1 and macro-averaged F1 scores

et al. 2021

View full text Add to dashboard Cite

A binary classification problem is common in medical field, and we often use sensitivity, specificity, accuracy, negative and positive predictive values as measures of performance of a binary predictor. In computer science, a classifier is usually evaluated with precision (positive predictive value) and recall (sensitivity). As a single summary measure of a classifier’s performance, F1 score, defined as the harmonic mean of precision and recall, is widely used in the context of information retrieval and information extraction evaluation since it possesses favorable characteristics, especially when the prevalence is low. Some statistical methods for inference have been developed for the F1 score in binary classification problems; however, they have not been extended to the problem of multi-class classification. There are three types of F1 scores, and statistical properties of these F1 scores have hardly ever been discussed. We propose methods based on the large sample multivariate central limit theorem for estimating F1 scores with confidence intervals.

show abstract

“…The lower the standard deviation indicates, the more near the average. Those features with the same standard deviation as the label are most helpful (H. Zhou, Ma, & Li, 2021).…”

Section: Feature Selectionmentioning

confidence: 99%

Time-Series Forecasting of Cryptocurrency Prices Using High-Dimensional Features and a Hybrid Approach

Joojili¹

View full text Add to dashboard Cite

Nowadays, digital cryptocurrencies are the most popular asset, especially for international exchanges. Bitcoin is the earliest cryptocurrency that succeeded in being used in financial transactions. Bitcoin stores the transactions in Blockchain technology. Bitcoin price has been unstable during the time from 0.5$ to about 60,000$ since 2010. Many efforts exist to predict Bitcoin value or its fluctuations using machine learning techniques. The price prediction is usually more challenging than fluctuations prediction, and its performance metrics are improved. This study introduces a methodology to predict Bitcoin price in a dataset, including four intervals to evaluate the proposed method in different situations. The experimental results show that the generalized linear model and Long Short-Term Memory (LSTM) were the best machine learning techniques. The proposed model outperforms the deep learning baseline model with about 18% and 20% relative improvement in mean absolute error and means absolute percentage error, respectively. Deep learning approaches have achieved much better results than other approaches due to the automatic selection of features. Compared to the results reported in the literature, the 1D-CNN+IndRNN proposed approach has reached 81% accuracy, with an 18% improvement. In the proposed approach, 1D-CNN is responsible for feature extraction and IndRNN is responsible for learning features in the form of time series.Appreciating to many people, from all over the world, who so generously helped me during my master's degree and contributed to this thesis work.Special mention goes to my supervisors, Professor Richard Yu and Professor Omair Shafiq. During my master's degree, they have always been providing their tremendous academic support. They are enthusiastic about academics, and their attitude encourages me a lot as a researcher working in data science. They also generously provided me funding help to support my family and my tuition fee, especially when I found out my mother got cancer and my situation was so difficult. I am also hugely appreciative to Professor David Thue and Dr. Gerry Chan, who encourage me to improve my academics.Last but not least, thanks to my family and friends, especially my mom and my wife. Thanks for supporting me to complete my master's degree abroad. This thesis work will never exist without their, not only because of financial and mental support, but also their passion for and put their belief in me. I wish my thesis work can help their as a return, even little. Thanks to all her generous support given to me all the time.

show abstract

Feature selection based on term frequency deviation rate for text classification

Cited by 13 publications

References 31 publications

A difficulty-aware and task-augmentation method based on meta-learning model for few-shot diabetic retinopathy classification

A difficulty-aware and task-augmentation method based on meta-learning model for few-shot diabetic retinopathy classification

Confidence interval for micro-averaged F1 and macro-averaged F1 scores

Time-Series Forecasting of Cryptocurrency Prices Using High-Dimensional Features and a Hybrid Approach

Contact Info

Product

Resources

About