Pith. sign in

REVIEW 4 major objections 6 minor 72 references

ED-Filter: Dynamic Feature Filtering for Eating Disorder Classification

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that ED-Filter finds small feature subsets that classify eating-disorder tweets more accurately than standard filters and wrappers, and fast enough for streaming data.

desk verdict The paper's central accuracy claims are uninterpretable because no held-out split is described, and the theoretical bound looks like a garbled Fano inequality. read the letter →

arxiv 2501.14785 v1 pith:SLUL7HCA submitted 2025-01-04 stat.ML cs.AIcs.LGcs.SI

Mehdi Naseriparsa , Suku Sukunesan , Zhen Cai , Osama Alfarraj , Amr Tolba , Saba Fathi Rabooki , Feng Xia This is my paper · ORCID
classification stat.MLcs.AIcs.LGcs.SI
keywords eatingdisorderdetectionfeatureselectionbranchandboundinformationgaingreedysearchdeeplearningTwitteranalysisclassification
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that a feature-selection method called ED-Filter, which ranks eating-disorder-related Twitter keywords by information gain and then searches the feature space with a branch-and-bound procedure, can find a small feature subset that classifies eating-disorder-related Twitter users more accurately than standard filters and wrappers. This matters because Twitter eating-disorder data is high-dimensional and changes frequently, so the authors argue that a fast, adaptive feature-selection step is needed to turn raw tweets into reliable classification models. They report that on a corpus of Pro-ED tweets, ED-Filter keeps precision around 82-87% as the dataset grows from 1,000 to 8,000 users, while the comparison filters drop below 75%, and that a hybrid greedy deep-learning variant achieves comparable accuracy in under 90 seconds for up to 11 features.

What carries the argument

The load-bearing object is the upper-bound estimate $\bar{\theta}(F,Y) = \frac{IG(Y;F) - \log(n) + 1}{\log(n-1)} + 1$ from Equation 7, which the paper derives by relating classification accuracy to information gain and binary entropy and claims is admissible, meaning it never underestimates the true accuracy. The branch-and-bound search keeps partial feature subsets in a max-heap, expanding the most promising entry and stopping when its upper bound falls below the best accuracy found so far. The hybrid greedy deep-learning variant restricts the search to a seed set of high-scoring features and to subsets whose size is predicted by a two-hidden-layer MLP trained on the Twitter data.

What would settle it

Re-run the ED-Filter comparisons with a strict temporal or random split: rank features and train the MLP cardinality predictor only on a training portion, search for the best feature subset there, and report classification accuracy on a held-out test portion. If ED-Filter's advantage over InfoGain, Relief, and the wrappers disappears or reverses on the held-out test set, the central claim is falsified.

Watch

Extended reading notes

Core claim

The central claim is that the optimal feature subset for eating-disorder classification can be identified by an informed branch-and-bound search whose pruning is driven by an admissible upper bound on accuracy derived from information gain, and that when the full search is too slow for streaming Twitter data, a greedy expansion restricted to seed features plus an MLP that predicts subset cardinality finds a near-optimal subset quickly. On the collected Twitter data, the authors report that ED-Filter achieves the best precision among InfoGain-Filter, Relief-Filter, Wrapper-BestFirst, and Wrapper-Forward, and that it selects fewer features than the compared methods. The hybrid method is presented as comparable to the greedy method, with accuracy differences below 5 percent on average, while halving or eliminating the combinatorial search overhead.

Load-bearing premise

The load-bearing premise is that the classification accuracy $\theta(F',Y)$ used to compare feature subsets is an honest measure of out-of-sample performance; the paper reports accuracy computed without describing a train/test split, so if the same records are used to select features and to measure accuracy, the reported gains could be an artifact of fitting the data.

Editorial extensions

If this is right

  • On the reported Twitter data, ED-Filter is the only method whose precision stays above 82% across data sizes from 1,000 to 8,000 users, while InfoGain and Relief filters fall below 75%.
  • ED-Filter selects fewer features than the compared filters and wrappers, which the paper argues makes it more scalable for dynamic Twitter streams.
  • The hybrid greedy deep-learning variant runs within about 90 seconds for feature sets up to 11 features, whereas the full branch-and-bound search time grows sharply beyond 6 features.
  • Accuracy differences between the hybrid and greedy methods stay below 5 percentage points on average, so the speed-up is claimed to come at a small effectiveness cost.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the reported accuracy is computed on the same records used to rank features and train the cardinality predictor, the gains may shrink under a proper temporal or random split; a reader should test this before relying on the headline numbers.
  • The upper-bound identity in Equation 7 is the kind of quantity that could be reused in other high-dimensional text classification tasks, but its admissibility would have to be re-derived for each new accuracy measure and data distribution.
  • Because the MLP cardinality predictor is trained on Twitter eating-disorder data, applying the hybrid method to a new platform or a new disorder would require re-training; the paper does not specify how often the model must be updated as the stream evolves.
  • The greedy add/remove local search depends on the order in which seed features are considered; testing multiple random seeds would reveal how stable the selected subset is to feature ordering.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript proposes ED-Filter, a feature-selection method for classifying eating-disorder-related Twitter users. The pipeline ranks 25 keyword-count features by information gain, searches for a high-accuracy subset with an informed branch-and-bound procedure whose pruning uses an information-theoretic upper bound (Eq. 7), and augments the search with a greedy variant and an MLP that predicts the desired subset cardinality. Experiments on a Twitter dataset of 11,620 user accounts with 37,405 tweets compare ED-Filter with InfoGain, Relief, and two wrapper baselines, and the paper reports that ED-Filter achieves the best precision and selects the fewest features. The abstract claims that ED-Filter iteratively identifies an optimal set of promising features and yields significant improvements in classification accuracy.

Significance. Social-media-based eating-disorder detection is a timely and practically relevant problem, and the paper addresses a real difficulty: high-dimensional, sparse keyword-count features. The proposed combination of filter-based ranking and bounded search is a reasonable design idea, and the authors provide a data-collection description, a public code repository, and comparisons against standard baselines. However, the central claims are not currently supported: the evaluation does not separate feature selection from final accuracy measurement, the upper-bound derivation in Section 5 is mathematically invalid, and the algorithm pseudocode is internally inconsistent. If these issues were corrected with a proper held-out evaluation and a valid admissibility proof, the work could make a useful contribution; as presented, the experimental and theoretical evidence is not sufficient.

major comments (4)
  1. [Section 8.2-8.3, Figures 3-4] The paper does not describe any train/test split or cross-validation for the reported accuracy. The same Twitter data are used to rank features by information gain (Section 4.1), to train the MLP cardinality predictor on data chunks (Section 7), to search for a feature subset, and to compute the final theta values in Figures 3 and 4. With selection and evaluation on the same records, the accuracy figures are optimistically biased and the claimed significant improvements over InfoGain-Filter, Relief-Filter, Wrapper-BestFirst, and Wrapper-Forward cannot be interpreted. A held-out evaluation, for example nested cross-validation in which feature ranking, cardinality prediction, and subset search are performed only on training folds, is required before any effectiveness claim can be made.
  2. [Section 5, Eq. (6)-(7), Lemma 1] The derivation of Lemma 1 is not valid. In Eq. (6), n is the number of features in F, but the entropy term log(n) in a mutual-information bound must refer to the cardinality of the class variable Y, since I(Y;F) is bounded by H(Y), which is at most log|Y|; using the feature count in this role mixes two different quantities. Algebraically, solving Eq. (6) for theta gives a bound of the form theta <= 1 - (log n - 1 - IG)/log(n-1), not the expression in Eq. (7), and the proof's step of replacing H2(theta) by 1 and adding 1 to the denominator changes the inequality. Because Lemma 2's admissibility argument depends entirely on this upper bound, the optimality claim for Algorithm 1 is unsupported.
  3. [Section 5, Algorithm 1; Section 7, Algorithm 2] In both algorithms, the measured accuracy is immediately overwritten by the upper bound: line 4 of Algorithm 1 sets e.theta = theta(e.F,Y) and line 5 sets e.theta to the Eq. (7) expression, so the heap ordering and the termination test e.theta < thetamin use only the bound and never the measured accuracy. This makes the behavior of the informed branch-and-bound procedure undefined as written. In addition, Eq. (2) defines theta(F',Y) as TP/(TP+FP), which is precision, not accuracy; the text and figures refer to this quantity as "classification accuracy" throughout Section 8, so the reported numbers need to be relabeled or the metric redefined.
  4. [Section 8.3, Table 3] The SVD/LDA comparison reports per-class precision values but gives no indication of how these are computed, whether the classes are balanced in the test data, or whether the same evaluation split is used as in Figures 3 and 4. The discussion attributes the pattern to class sizes, but no class-size statistics are reported, so the comparison cannot be independently assessed.
minor comments (6)
  1. [Section 1] The text says "we propose FilterBoost" in the paragraph beginning "In this paper, we propose FilterBoost," but the method is everywhere else called ED-Filter; this inconsistent name should be reconciled.
  2. [Section 8.1] The data-processing description does not explain how the 37,405 tweets are mapped to the 11,620 user accounts or how the four class labels y in {0,1,2,3} are assigned; without this, the construction of the classification rows in Table 1 is incomplete.
  3. [Figure 3(e)-(f)] The terms "worst-case scenario" and "best-case scenario" are not defined, and no number of runs, random seeds, or error bars are given despite the text saying each method was run "a couple of times."
  4. [Section 8.4, Figure 6] The information-gain threshold values that yield 25, 18, 14, and 11 eligible features are never reported, which makes the scalability experiments difficult to reproduce.
  5. [Algorithms 1 and 2] The getNext(F) calls iterate over the full feature set; the algorithms should state explicitly that only features not already in e.F are considered and that the loops terminate when no such feature remains.
  6. [References] There are spelling inconsistencies in author names, for example "Dodzilo" in the text versus "Dondzilo" in references [53] and [54], and "Sedghir" in the text versus "Seghir" in reference [62]; these should be corrected.

Circularity Check

2 steps flagged · score 8.0 of 10

Reported accuracy is the same objective the method optimizes: ED-Filter selects and sizes feature subsets by maximizing θ on the Twitter data, then reports θ on that same data as effectiveness, so the claimed gains are in-sample by construction.

  1. fitted input called prediction [Section 7 (The Hybrid Solution), paragraph on building the cardinality detector; Algorithm 2 line 3]
    "To find the optimal feature subset size |F ∗|, we run a comprehensive search on the feature space Fi of each Twitter data chunk Ti ⊂ T and compute the classification accuracy θ(Fi, Yi) for the Twitter data on eating disorder types. Then, we choose the feature subset cardinality that has led to the best classification accuracy to feed the neural network for training purposes."

    The MLP 'predicts' the feature subset cardinality c, but c is defined as the cardinality that achieved the best accuracy θ on chunks of the same Twitter dataset T. Algorithm 2 then constrains the search to that fitted cardinality and the experiments report accuracy on the same Twitter data. The reported effectiveness of the hybrid method is therefore the value of the very objective used to fit the cardinality predictor, not an independent out-of-sample result.

  2. self definitional [Section 4.2 (Iterative Feature Reduction); Section 8.2 (Effectiveness Assessment)]
    "We assess the classification performance of a feature set to determine the irrelevant features in each round. For example, after removing F ′ from F , we assess the eating disorder classification model on the new feature set θ(F ′′, Y). If the performance improves (this means that the removed features were irrelevant), we update the feature set with the newly reduced feature set. ... We use the classification accuracy of the reduced (filtered) eating disorder data to verify the effectiveness of the feature selection techniques."

    The selection criterion of ED-Filter is θ(F,Y) itself: a feature subset is kept or discarded depending on whether accuracy on the Twitter data improves. Section 8 then reports classification accuracy on the same Twitter data as evidence of effectiveness, and no train/test split is described. The reported accuracy is thus the optimized objective, not a prediction; the claimed 'significant improvements' over InfoGain, Relief, and wrappers are not independent of the data used to choose the subset.

full rationale

The paper's branch-and-bound proof and Equation 7 are not themselves circular: Lemma 1 attempts to bound accuracy by mutual information, and Lemma 2 is a standard admissibility argument if the bound is valid. The circularity is in the empirical validation chain. Both the feature subset and the cardinality of the subset are selected by maximizing θ on the same Twitter dataset whose classification accuracy is later reported as the method's effectiveness. Section 7 explicitly says the optimal cardinality is found by searching each Twitter data chunk and choosing the cardinality that led to the best classification accuracy, which is then fed to the MLP as the training target. Algorithm 2 uses that fitted cardinality to prune the search, and Section 8 reports accuracy on the same data without any described held-out split. Consequently, the abstract's central claim of 'significant improvements in classification accuracy' reduces to reporting the value of the objective that was optimized by construction. The paper's self-citations, such as the authors' earlier EDBase work, are not load-bearing for this claim, so the score reflects the evaluation-loop circularity rather than citation practice.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central claim rests on an unvalidated evaluation protocol, a dubious information-theoretic bound, an unsound pruning rule, and a neural network trained on the same data used for evaluation. These are the main things the reader must accept without independent support.

free parameters (5)
  • MLP cardinality predictor hyperparameters = unspecified
    Layer sizes, epochs, learning rate, batch size, and validation protocol are not reported (Section 7).
  • Information gain threshold = unspecified
    Controls how many of the 25 features enter the search; values vary in Figure 6 without a selection criterion (Section 8.4).
  • Iteration limit = unspecified
    Described as a tuning parameter affecting dynamic performance (Section 4), no default or tuning strategy given.
  • Number of CorEx topics = unspecified
    Chosen among 2-50 as the 'best model' without defining the criterion (Section 8.1).
  • Seed feature set size = unspecified
    Seed features are chosen by single-feature accuracy (Section 6), but the number of seeds and the accuracy threshold are unspecified.
assumptions (6)
  • standard math Fano-type inequality relating accuracy and information gain (Eq 6)
    Underlies Lemma 1; valid only for uniform class distribution with n as the number of classes, not the 25 features used in the algorithms.
  • ad hoc to paper The Eq 7 upper bound is admissible for all feature subsets
    Admissibility is asserted in Lemma 2; the derivation replaces binary entropy with its maximum and applies the bound with the wrong n, so no valid guarantee is established.
  • ad hoc to paper Partial-subset accuracies can serve as lower bounds for pruning
    Algorithm 1 updates θmin with e.θ for any expanded node, but Property 2 states θ is non-monotone, so a partial subset's accuracy does not bound its descendants.
  • domain assumption Reported accuracies are unbiased out-of-sample estimates
    The paper never describes a held-out split for the accuracy measurements in Sections 8.2-8.3; selection and evaluation appear to use the same data.
  • domain assumption Multinomial Naive Bayes is a suitable evaluation classifier
    Used to compute θ for all methods (Section 8.3) without justification or comparison to other classifiers.
  • domain assumption The 25 expert-defined features and 4 classes represent the true ED classification task
    Features come from CorEx topics and expert consultation (Section 8.1); no validation of labels or feature definitions is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ED-Filter: Dynamic Feature Filtering for Eating Disorder Classification." pith.science (2026). https://pith.science/paper/SLUL7HCA

@misc{pith2026250114785,
  author       = {Pith},
  title        = {Pith review of: ED-Filter: Dynamic Feature Filtering for Eating Disorder Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SLUL7HCA}},
  note         = {Machine review of arXiv:2501.14785}
}
read the original abstract

Eating disorders (ED) are critical psychiatric problems that have alarmed the mental health community. Mental health professionals are increasingly recognizing the utility of data derived from social media platforms such as Twitter. However, high dimensionality and extensive feature sets of Twitter data present remarkable challenges for ED classification. To overcome these hurdles, we introduce a novel method, an informed branch and bound search technique known as ED-Filter. This strategy significantly improves the drawbacks of conventional feature selection algorithms such as filters and wrappers. ED-Filter iteratively identifies an optimal set of promising features that maximize the eating disorder classification accuracy. In order to adapt to the dynamic nature of Twitter ED data, we enhance the ED-Filter with a hybrid greedy-based deep learning algorithm. This algorithm swiftly identifies sub-optimal features to accommodate the ever-evolving data landscape. Experimental results on Twitter eating disorder data affirm the effectiveness and efficiency of ED-Filter. The method demonstrates significant improvements in classification accuracy and proves its value in eating disorder detection on social media platforms.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

72 extracted references · 72 canonical work pages

  1. [1]

    Risk factors for eating disorders: findings from a rapid review

    Sarah Barakat, Siˆ an A McLean, Emma Bryant, Anvi Le, Peta Marks, Stephen Touyz, and Sarah Maguire. Risk factors for eating disorders: findings from a rapid review. Journal of Eating Disorders , 11(1):8, 2023

    work page 2023

  2. [2]

    A mixed- studies systematic review of the experiences of body image, disordered eating, and eating disorders during the covid-19 pandemic

    Jekaterina Schneider, Georgina Pegram, Benjamin Gibson, Deborah Talamonti, Aline Tinoco, Nadia Craddock, Emily Matheson, and Mark Forshaw. A mixed- studies systematic review of the experiences of body image, disordered eating, and eating disorders during the covid-19 pandemic. International Journal of Eating Disorders, 56(1):26–67, 2023

    work page 2023

  3. [3]

    Diagnosis and prognosis of mental disorders by means of eeg and deep learning: a systematic mapping study

    Manuel J Rivera, Miguel A Teruel, Alejandro Mate, and Juan Trujillo. Diagnosis and prognosis of mental disorders by means of eeg and deep learning: a systematic mapping study. Artificial Intelligence Review , pages 1–43, 2022

    work page 2022

  4. [4]

    Eating and exercise behaviors in eating disorders and the general population during the covid-19 pandemic in australia: Initial results from the collate project

    Andrea Phillipou, Denny Meyer, Erica Neill, Eric J Tan, Wei Lin Toh, Tamsyn E Van Rheenen, and Susan L Rossell. Eating and exercise behaviors in eating disorders and the general population during the covid-19 pandemic in australia: Initial results from the collate project. International Journal of Eating Disorders , 53(7):1158–1165, 2020. 24

    work page 2020

  5. [5]

    Pike, Stuart B

    Deborah Mitchison, Jonathan Mond, Kay Bussey, Scott Griffiths, Nora Trompeter, Alexandra Lonergan, Kathleen M. Pike, Stuart B. Murray, and Phillipa Hay. Dsm-5 full syndrome, other specified, and unspecified eating disor- ders in australian adolescents: prevalence and clinical significance. Psychological Medicine, 2019

    work page 2019

  6. [6]

    The neurobiology of eating disorders

    Guido KW Frank, Megan E Shott, and Marisa C DeGuzman. The neurobiology of eating disorders. Child and Adolescent Psychiatric Clinics , 28(4):629–640, 2019

    work page 2019

  7. [7]

    Predicting mental health problems with personality, behavior, and social networks

    Dongyu Zhang, Teng Guo, Shiyu Han, Sadaf Vahabli, Mehdi Naseriparsa, and Feng Xia. Predicting mental health problems with personality, behavior, and social networks. In 2021 IEEE International Conference on Big Data (Big Data) , pages 4537–4546. IEEE, 2021

    work page 2021

  8. [8]

    Alina Arseniev-Koehler, Hedwig Lee, Tyler McCormick, and Megan A. Moreno. #proana: Pro-eating disorder socialization on twitter. Journal of Adolescent Health, 58(6):659–664, 2016

    work page 2016

Show all 72 references
  1. [9]

    The impact of covid-19 lockdown on social network sites use, body image disturbances and self-esteem among adolescent and young women

    Helena Vall-Roqu´ e, Ana Andr´ es, and Carmina Salda˜ na. The impact of covid-19 lockdown on social network sites use, body image disturbances and self-esteem among adolescent and young women. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 110:110293, 2021

    2021

  2. [10]

    A systematic review of the impact of the use of social networking sites on body image and disordered eating outcomes

    Grace Holland and Marika Tiggemann. A systematic review of the impact of the use of social networking sites on body image and disordered eating outcomes. Body Image, 17:100–110, 2016

    2016

  3. [11]

    Twitter: a good place to detect health conditions

    V ´ ıctor M Prieto, Sergio Matos, Manuel Alvarez, Fidel Cacheda, and Jos´ e Lu ´ ıs Oliveira. Twitter: a good place to detect health conditions. PloS one, 9(1):e86191, 2014

    2014

  4. [12]

    Data-driven computational social science: A survey

    Jun Zhang, Wei Wang, Feng Xia, Yu-Ru Lin, and Hanghang Tong. Data-driven computational social science: A survey. Big Data Research, 21:100145, 2020

    2020

  5. [13]

    Mobile multimedia recommendation in smart communities: A survey.IEEE Access, 1:606–624, 2013

    Feng Xia, Nana Yaw Asabere, Ahmedin Mohammed Ahmed, Jing Li, and Xiangjie Kong. Mobile multimedia recommendation in smart communities: A survey.IEEE Access, 1:606–624, 2013

    2013

  6. [14]

    Classification of twitter users with eating disorder engagement: Learning from the biographies

    Mohammad Abuhassan, Tarique Anwar, Matthew Fuller-Tyszkiewicz, Hannah K Jarman, Adrian Shatte, Chengfei Liu, and Suku Sukunesan. Classification of twitter users with eating disorder engagement: Learning from the biographies. Computers in Human Behavior , 140:107519, 2023

    2023

  7. [15]

    Jarman, Mohammad Abuhassan, Adrian Shatte, WIRED Team, and Suku Sukunesan

    Tarique Anwar, Matthew Fuller-Tyszkiewicz, Hannah K. Jarman, Mohammad Abuhassan, Adrian Shatte, WIRED Team, and Suku Sukunesan. Edbase: Gen- erating a lexicon base for eating disorders via social media. IEEE J. Biomed. Health Informatics, 26(12):6116–6125, 2022. 25

    2022

  8. [16]

    Haynos, and Rui Zhang

    Sicheng Zhou, Yunpeng Zhao, Rubina Rizvi, Jiang Bian, Ann F. Haynos, and Rui Zhang. Analysis of twitter to identify topics related to eating disorder symptoms. In 2019 IEEE International Conference on Healthcare Informatics (ICHI) , pages 1–4, 2019

    2019

  9. [17]

    Tweeting weight loss: A comparison of #thinspiration and #fitspiration communities on twitter

    Marika Tiggemann, Owen Churches, Lewis Mitchell, and Zoe Brown. Tweeting weight loss: A comparison of #thinspiration and #fitspiration communities on twitter. Body Image, 25:133–138, 2018

    2018

  10. [18]

    Clear, Eric Gilbert, and Munmun De Choudhury

    Stevie Chancellor, Jessica Annette Pater, Trustin A. Clear, Eric Gilbert, and Munmun De Choudhury. #thyghgapp: Instagram content moderation and lex- ical variation in pro-eating disorder communities. In Proceedings of the 19th ACM Conference on Computer-Supported Cooperative W...

    2016

  11. [19]

    Fitspiration and thinspiration: a comparison across three social networking sites

    Angela S Alberga, Samantha J Withnell, and Kristin M von Ranson. Fitspiration and thinspiration: a comparison across three social networking sites. Journal of Eating Disorders, 6(1):1–10, 2018

    2018

  12. [20]

    A review of feature selection methods with applications

    Alan Jovic, Karla Brkic, and Nikola Bogunovic. A review of feature selection methods with applications. In 38th International Convention on Information and Communication Technology, Electronics and Microelectronics, MIPRO 2015, Opatija, Croatia, May 25-29, 2015 , pages 1200–12...

    2015

  13. [21]

    Shifu2: A network representation learning based model for advisor-advisee relationship mining

    Jiaying Liu, Feng Xia, Lei Wang, Bo Xu, Xiangjie Kong, Hanghang Tong, and Irwin King. Shifu2: A network representation learning based model for advisor-advisee relationship mining. IEEE Transactions on Knowledge and Data Engineering, 33(4):1763–1777, 2019

    2019

  14. [22]

    Feature subset selection for multi-scale neighborhood decision information system via mutual information

    Lujing Zhang, Guoping Lin, Ling Wei, and Yi Kou. Feature subset selection for multi-scale neighborhood decision information system via mutual information. Artificial Intelligence Review , 57(1):1–30, 2024

    2024

  15. [23]

    Deep outdated fact detection in knowledge graphs

    Huiling Tu, Shuo Yu, Vidya Saikrishna, Feng Xia, and Karin Verspoor. Deep outdated fact detection in knowledge graphs. In IEEE International Conference on Data Mining Workshops (ICDMW) , pages 1443–1452, Shanghai, China, 2023. IEEE

    2023

  16. [24]

    A review of artificial intelligence methods for engineering prognostics and health management with implementation guidelines

    Khanh TP Nguyen, Kamal Medjaher, and Do T Tran. A review of artificial intelligence methods for engineering prognostics and health management with implementation guidelines. Artificial Intelligence Review, 56(4):3659–3709, 2023

    2023

  17. [25]

    A correlation- based feature weighting filter for naive bayes

    Liangxiao Jiang, Lungan Zhang, Chaoqun Li, and Jia Wu. A correlation- based feature weighting filter for naive bayes. IEEE Trans. Knowl. Data Eng. , 31(2):201–213, 2019. 26

    2019

  18. [26]

    Vallejo, Jos´ e A

    Carlos G. Vallejo, Jos´ e A. Troyano, Fernando Enr ´ ıquez, F. Javier Ortega, and Ferm ´ ın L. Cruz. MCFS: min-cut-based feature-selection. Knowl. Based Syst. , 195:105604, 2020

    2020

  19. [27]

    Data-driven approximations to np-hard problems

    Anton Milan, S Rezatofighi, Ravi Garg, Anthony Dick, and Ian Reid. Data-driven approximations to np-hard problems. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 31, 2017

    2017

  20. [28]

    A framework for clustering and dynamic maintenance of XML documents

    Ahmed Al-Shammari, Chengfei Liu, Mehdi Naseriparsa, Bao Quoc Vo, Tarique Anwar, and Rui Zhou. A framework for clustering and dynamic maintenance of XML documents. In Advanced Data Mining and Applications - 13th International Conference, ADMA 2017, Singapore, November 5-6, 2017...

    2017

  21. [29]

    Saiful Islam, Chengfei Liu, and Irene Moser

    Mehdi Naseriparsa, Md. Saiful Islam, Chengfei Liu, and Irene Moser. No-but- semantic-match: computing semantically matched xml keyword search results. World Wide Web , 21(5):1223–1257, 2018

    2018

  22. [30]

    Saiful Islam, Chengfei Liu, and Lu Chen

    Mehdi Naseriparsa, Md. Saiful Islam, Chengfei Liu, and Lu Chen. Xsnippets: Exploring semi-structured data via snippets. Data Knowl. Eng. , 124, 2019

    2019

  23. [31]

    Saiful Islam, and Rui Zhou

    Mehdi Naseriparsa, Chengfei Liu, Md. Saiful Islam, and Rui Zhou. Xplorerank: exploring XML data via you may also like queries. World Wide Web , 22(4):1727– 1750, 2019

    2019

  24. [32]

    A dynamic mlp-based ddos attack detection method using feature selection and feedback

    Meng Wang, Yiqin Lu, and Jiancheng Qin. A dynamic mlp-based ddos attack detection method using feature selection and feedback. Computers & Security , 88:101645, 2020

    2020

  25. [33]

    Semi- supervised feature selection via hierarchical regression for web image classifica- tion

    Xiaonan Song, Jianguang Zhang, Yahong Han, and Jianmin Jiang. Semi- supervised feature selection via hierarchical regression for web image classifica- tion. Multim. Syst. , 22(1):41–49, 2016

    2016

  26. [34]

    Hira and Duncan Fyfe Gillies

    Zena M. Hira and Duncan Fyfe Gillies. A review of feature selection and feature extraction methods applied on microarray data. Adv. Bioinformatics , 2015:198363:1–198363:13, 2015

    2015

  27. [35]

    A framework for processing cumulative frequency queries over medical data streams

    Ahmed Al-Shammari, Rui Zhou, Chengfei Liu, Mehdi Naseriparsa, and Bao Quoc Vo. A framework for processing cumulative frequency queries over medical data streams. In Web Information Systems Engineering - WISE 2018 - 19th International Conference, Dubai, United Arab Emirates, No...

    2018

  28. [36]

    Benchmark for filter methods for feature selection in high-dimensional classification data

    Andrea Bommert, Xudong Sun, Bernd Bischl, J¨ org Rahnenf¨ uhrer, and Michel Lang. Benchmark for filter methods for feature selection in high-dimensional classification data. Computational Statistics & Data Analysis , 143:106839, 2020. 27

    2020

  29. [37]

    Feature selection and classification–a probabilis- tic wrapper approach

    Huan Liu and Rudy Setiono. Feature selection and classification–a probabilis- tic wrapper approach. In Industrial and Engineering Applications or Artificial Intelligence and Expert Systems , pages 419–424. CRC Press, 2022

    2022

  30. [38]

    A new feature selec- tion technique for load and price forecast of electrical power systems

    Oveis Abedinia, Nima Amjady, and Hamidreza Zareipour. A new feature selec- tion technique for load and price forecast of electrical power systems. IEEE Transactions on Power Systems , 32(1):62–74, 2017

    2017

  31. [39]

    Toward optimal feature selection in naive bayes for text categorization

    Bo Tang, Steven Kay, and Haibo He. Toward optimal feature selection in naive bayes for text categorization. IEEE Trans. Knowl. Data Eng. , 28(9):2508–2521, 2016

    2016

  32. [40]

    Scstcf: spatial-channel selection and temporal regularized correlation filters for visual tracking

    Jianming Zhang, Wenjun Feng, Tingyu Yuan, Jin Wang, and Arun Kumar Sanga- iah. Scstcf: spatial-channel selection and temporal regularized correlation filters for visual tracking. Applied Soft Computing , 118:108485, 2022

    2022

  33. [41]

    Mctwo: a two-step feature selection algorithm based on maximal information coefficient

    Ruiquan Ge, Manli Zhou, Youxi Luo, Qinghan Meng, Guoqin Mai, Dongli Ma, Guoqing Wang, and Fengfeng Zhou. Mctwo: a two-step feature selection algorithm based on maximal information coefficient. BMC Bioinformatics , 17:142, 2016

    2016

  34. [42]

    Fea- ture selection using lebesgue and entropy measures for incomplete neighborhood decision systems

    Lin Sun, Lanying Wang, Yuhua Qian, Jiucheng Xu, and Shiguang Zhang. Fea- ture selection using lebesgue and entropy measures for incomplete neighborhood decision systems. Knowl. Based Syst. , 186, 2019

    2019

  35. [43]

    Machine learning for mental health in social media: Bibliometric study

    Jina Kim, Daeun Lee, and Eunil Park. Machine learning for mental health in social media: Bibliometric study. J Med Internet Res , 23(3):e24870, 2021

    2021

  36. [44]

    A framework for multi-faceted content analysis of social media chatter regarding non-medical use of prescription medications

    Shaina Raza, Brian Schwartz, Sahithi Lakamana, Yao Ge, and Abeed Sarker. A framework for multi-faceted content analysis of social media chatter regarding non-medical use of prescription medications. BMC Digit Health , 1(29), 2023

    2023

  37. [45]

    Mining social media for prescription medication abuse monitoring: a review and proposal for a data- centric framework

    Abeed Sarker, Annika DeRoos, and Jeanmarie Perrone. Mining social media for prescription medication abuse monitoring: a review and proposal for a data- centric framework. J. Am. Med. Inform. Assoc. , 27(2):315–329, 2020

    2020

  38. [46]

    Sentiment analysis in social media data for depression detection using artificial intelligence: A review

    Nirmal Varghese Babu and E Grace Mary Kanaga. Sentiment analysis in social media data for depression detection using artificial intelligence: A review. SN Comput. Sci. , 3(74), 2022

    2022

  39. [47]

    Toro, Mahmood Alzubaidi, Sara Irshaidat, Hashem Abu Serhan, Alaa A

    Arfan Ahmed, Sarah Aziz, Carla T. Toro, Mahmood Alzubaidi, Sara Irshaidat, Hashem Abu Serhan, Alaa A. Abd-alrazaq, and Mowafa Househ. Machine learn- ing models to detect anxiety and depression through social media: A scoping review. Computer Methods and Programs in Biomedicine...

    2022

  40. [48]

    Annisa Ristya Rahmanti, Chia-Hui Chien, Aldilas Achmad Nursetyo, Atina Hus- nayain, Bayu Satria Wiratama, Anis Fuad, Hsuan-Chia Yang, and Yu-Chuan Jack Li. Social media sentiment analysis to monitor the performance of vaccination 28 coverage during the early phase of the natio...

    2022

  41. [49]

    Predictors, mediators, and moderators of response to digital interventions for eating disorders: A systematic review

    Zoe McClure, Matthew Fuller-Tyszkiewicz, Mariel Messer, and Jake Linardon. Predictors, mediators, and moderators of response to digital interventions for eating disorders: A systematic review. International Journal of Eating Disorders , 2023

    2023

  42. [50]

    Bert model-based approach for detecting categories of tweets in the field of eating disorders (ed)

    Jos´ e Alberto Ben ´ ıtez-Andrades, Jos´ e Manuel Alija-P´ erez, Isa ´ ıas Garc ´ ıa- Rodr ´ ıguez, Carmen Benavides, H´ ector Alaiz-Moret´ on, Rafael Pastor Vargas, and Mar ´ ıa Teresa Garc ´ ıa-Ord´ as. Bert model-based approach for detecting categories of tweets in the fiel...

    2021

  43. [51]

    Prevalence of eating disorders over the 2000–2018 period: a systematic literature review

    Marie Galmiche, Pierre D´ echelotte, Gr´ egory Lambert, and Marie Pierre Tavolacci. Prevalence of eating disorders over the 2000–2018 period: a systematic literature review. The American Journal of Clinical Nutrition , 109(5):1402–1413, 2019

    2000

  44. [52]

    Concurrent and prospective analyses of peer, television and social media influ- ences on body dissatisfaction, eating disorder symptoms and life satisfaction in adolescent girls

    Christopher J Ferguson, M´ onica E Mu˜ noz, Adolfo Garza, and Mariza Galindo. Concurrent and prospective analyses of peer, television and social media influ- ences on body dissatisfaction, eating disorder symptoms and life satisfaction in adolescent girls. J Youth Adolesc, 43:...

    2014

  45. [53]

    Rodgers, and Fanny Alexandra Dietel

    Laura Dondzilo, Rachel F. Rodgers, and Fanny Alexandra Dietel. Association between engagement with appearance and eating related tiktok content and eat- ing disorder symptoms via recommended content and appearance comparisons. International Journal of Eating Disorders , 57(2):...

    2024

  46. [54]

    Laura Dondzilo, Tamsin Mahalingham, and Patrick J.F. Clarke. A preliminary investigation of the causal role of social media use in eating disorder symptoms. Journal of Behavior Therapy and Experimental Psychiatry , 82:101923, 2024

    2024

  47. [55]

    McLean Rachel F Rodgers and Susan J

    Siˆ an A. McLean Rachel F Rodgers and Susan J. Paxton. Enhancing understand- ing of social media literacy to better inform prevention of body image and eating disorders. Eating Disorders, 0(0):1–19, 2024

    2024

  48. [56]

    Whistleblowing and tech on twitter

    Laura Duits, Isha Kashyap, Joey Bekkink, Kousar Aslam, and Emitz´ a Guzm´ an. Whistleblowing and tech on twitter. In 2023 IEEE/ACM 20th International Conference on Mining Software Repositories (MSR) , pages 296–308, 2023

    2023

  49. [57]

    Empow- ering machine learning models with contextual knowledge for enhancing the detection of eating disorders in social media posts

    Jos´ e Alberto Ben ´ ıtez-Andrades, Mar ´ ıa Teresa Garc ´ ıa-Ord´ as, Mayra Russo, Ahmad Sakor, Luis Daniel Fernandes Rotger, and Maria-Esther Vidal. Empow- ering machine learning models with contextual knowledge for enhancing the detection of eating disorders in social media...

    2023

  50. [58]

    Methodologies for monitoring mental health on twitter: systematic review

    Nina H Di Cara, Valerio Maggio, Oliver SP Davis, and Claire MA Haworth. Methodologies for monitoring mental health on twitter: systematic review. Journal of Medical Internet Research , 25:e42734, 2023

    2023

  51. [59]

    A novel adaptive memetic binary optimization algorithm for feature selection

    Ahmet Cevahir Cinar. A novel adaptive memetic binary optimization algorithm for feature selection. Artificial Intelligence Review , pages 1–58, 2023

    2023

  52. [60]

    Gayathri Devi and M

    S. Gayathri Devi and M. Sabrigiriraj. Feature selection, online feature selection techniques for big data classification: - a review. In 2018 International Conference on Current Trends towards Converging Technologies (ICCTCT), pages 1–9, 2018

    2018

  53. [61]

    Laxmi Arya and Govind P. Gupta. Ensemble filter-based feature selection model for cyber attack detection in industrial internet of things. In 2023 9th International Conference on Advanced Computing and Communication Systems (ICACCS), volume 1, pages 834–840, 2023

    2023

  54. [62]

    Wrapper- based feature selection for medical diagnosis: The btlbo-knn algorithm

    Fateh Seghir, Ahlem Drif, Saadeddine Selmani, and Hocine Cherifi. Wrapper- based feature selection for medical diagnosis: The btlbo-knn algorithm. IEEE Access, 11:61368–61389, 2023

    2023

  55. [63]

    Logtiw:a log anomaly detection model based on tf-idf weighted semantic features

    Jia Kang, Junfeng Zhao, and Zhengxin Li. Logtiw:a log anomaly detection model based on tf-idf weighted semantic features. In2024 International Joint Conference on Neural Networks (IJCNN) , pages 1–7, 2024

    2024

  56. [64]

    A review of feature selection and its methods

    B Venkatesh and J Anuradha. A review of feature selection and its methods. Cybernetics and Information Technologies, 19(1):3–26, 2019

    2019

  57. [65]

    Discovering structure in high-dimensional data through correlation explanation

    Greg Ver Steeg and Aram Galstyan. Discovering structure in high-dimensional data through correlation explanation. In Advances in Neural Information Process- ing Systems 27: Annual Conference on Neural Information Processing Systems 2014, December 8-13 2014, Montreal, Quebec, C...

    2014

  58. [66]

    Rizvi, Yefeng Wang, Thao Nguyen, Jake Vasilakes, Jiang Bian, Zhe He, and Rui Zhang

    Rubina F. Rizvi, Yefeng Wang, Thao Nguyen, Jake Vasilakes, Jiang Bian, Zhe He, and Rui Zhang. Analyzing social media data to understand consumer information needs on dietary supplements. In MEDINFO 2019: Health and Wellbeing e- Networks for All - Proceedings of the 17th World ...

    2019

  59. [67]

    A feature selection method based on information gain and genetic algo- rithm

    Shang Lei. A feature selection method based on information gain and genetic algo- rithm. In 2012 International Conference on Computer Science and Electronics Engineering, volume 2, pages 355–358, 2012

    2012

  60. [68]

    Urbanowicz, Melissa Meeker, William La Cava, Randal S

    Ryan J. Urbanowicz, Melissa Meeker, William La Cava, Randal S. Olson, and Jason H. Moore. Relief-based feature selection: Introduction and review. Journal of Biomedical Informatics , 85:189–203, 2018. 30

    2018

  61. [69]

    Ron Kohavi and George H. John. Wrappers for feature subset selection. Artif. Intell., 97(1-2):273–324, 1997

    1997

  62. [70]

    G´ amez, and Jos´ e M

    Pablo Bermejo, Luis de la Ossa, Jos´ e A. G´ amez, and Jos´ e M. Puerta. Fast wrapper feature subset selection in high-dimensional datasets by means of filter re-ranking. Knowledge-Based Systems, 25(1):35–44, 2012

    2012

  63. [71]

    Competitive swarm optimized svd clutter filtering for ultrafast power doppler imaging

    Yinran Chen, Baohui Fang, Fengling Meng, Jianwen Luo, and Xiongbiao Luo. Competitive swarm optimized svd clutter filtering for ultrafast power doppler imaging. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 71(4):459–473, 2024

    2024

  64. [72]

    A revised formation of trace ratio lda for small sample size problem

    Zhengxin Li, Feiping Nie, Rong Wang, and Xuelong Li. A revised formation of trace ratio lda for small sample size problem. IEEE Transactions on Neural Networks and Learning Systems , 2024. 31

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.