FS Method Research Articles

Feature selection through validation and un-censoring of endovascular repair survival data for predicting the risk of re-intervention

BackgroundFeature selection (FS) process is essential in the medical area as it reduces the effort and time needed for physicians to measure unnecessary features. Choosing useful variables is a difficult task with the presence of censoring which is the unique characteristic in survival analysis. Most survival FS methods depend on Cox’s proportional hazard model; however, machine learning techniques (MLT) are preferred but not commonly used due to censoring. Techniques that have been proposed to adopt MLT to perform FS with survival data cannot be used with the high level of censoring. The researcher’s previous publications proposed a technique to deal with the high level of censoring. It also used existing FS techniques to reduce dataset dimension. However, in this paper a new FS technique was proposed and combined with feature transformation and the proposed uncensoring approaches to select a reduced set of features and produce a stable predictive model.MethodsIn this paper, a FS technique based on artificial neural network (ANN) MLT is proposed to deal with highly censored Endovascular Aortic Repair (EVAR). Survival data EVAR datasets were collected during 2004 to 2010 from two vascular centers in order to produce a final stable model. They contain almost 91% of censored patients. The proposed approach used a wrapper FS method with ANN to select a reduced subset of features that predict the risk of EVAR re-intervention after 5 years to patients from two different centers located in the United Kingdom, to allow it to be potentially applied to cross-centers predictions. The proposed model is compared with the two popular FS techniques; Akaike and Bayesian information criteria (AIC, BIC) that are used with Cox’s model.ResultsThe final model outperforms other methods in distinguishing the high and low risk groups; as they both have concordance index and estimated AUC better than the Cox’s model based on AIC, BIC, Lasso, and SCAD approaches. These models have p-values lower than 0.05, meaning that patients with different risk groups can be separated significantly and those who would need re-intervention can be correctly predicted.ConclusionThe proposed approach will save time and effort made by physicians to collect unnecessary variables. The final reduced model was able to predict the long-term risk of aortic complications after EVAR. This predictive model can help clinicians decide patients’ future observation plan.

Evaluation of an enhanced FS method for finding the initiation time of earthquake-induced landslides

In this paper, an examination of an enhanced FS method used for finding the initiation time of earthquake-induced landslides is presented. Validation was conducted via comparisons between the predicted and reported results of a benchmark site subjected to two major earthquake events: at Chi-Chi in 1999 (7.6 Mw) and Meinong in 2016 (6.4 Mw). To further validate the applicability of the ground motions utilized in this study, a finite element analysis was also carried out. Based on the results, satisfactory agreement between the predictions obtained by the enhanced FS method and the observed results in terms of the landslide initiation time was realized. Conclusions drawn from the study show that (a) the enhanced FS method can be both efficient and practical in finding the likely initiation time of a landslide, and it was found that the initiation time of the Tsaoling landslide during the Chi-Chi earthquake lay between 37.5 and 38.5 s; (b) when validating the enhanced FS method, the friction angle of 38.5o obtained from direct shear tests rather than a smaller value of 20.8o obtained from rotary-shear high-velocity friction experiments, led to a more appropriate results; and (c) the failure surface obtained by finite element codes was satisfactory when compared to the observed result.

Arabic Sentiment Analysis with Optimal Combination of Features Selection and Machine Learning Approaches

The main objective of this research study is to design a model that allows for the utilization of a novel technique for the implementation of sentiment analysis in the Arabic language. Sentiment analysis is an interesting task that includes web mining, Natural Language Processing (NLP) and Machine Learning (ML). Most of the research work on sentiment analysis was focused on the texts in the English language. Therefore, the research on sentiment analysis in the Arabic language and other languages are in the infancy stage. This study empirically evaluates three Feature Selection Methods (FSM) (Information Gain (IG), Chi-square (CHI) and Gini Index (GI)) and, three classification approaches (Association Rule (AR) mining and the N-gram model and the Meta-classifier approach) for the implementation of sentiment classification in the Arabic language. A number of related experiments have been carried out on the Opinion Corpus of Arabic (OCA). The results obtained from the experiments were favorable, depending on the algorithms used and the number of selected feature has proven that the use of FS method can increase the performance of sentiment classification in the Arabic language. The results of the experiments reveal that FS method is obtained to develop the classifier performance. Furthermore, the results of the experiment indicated that the use of CHI feature selection has produced the best performance for FS and the performance of meta-classifier a combination approach has outperformed the other approaches for sentiment classification in the Arabic language. In conclusion, this research study has proven that the combination approach (meta-classifier) with the chi-square FS method produces the most accurate classification technique, as high as 90.80%.

Estimation of Discriminative Feature Subset Using Community Modularity.

Feature selection (FS) is an important preprocessing step in machine learning and data mining. In this paper, a new feature subset evaluation method is proposed by constructing a sample graph (SG) in different k-features and applying community modularity to select highly informative features as a group. However, these features may not be relevant as an individual. Furthermore, relevant in-dependency rather than irrelevant redundancy among the selected features is effectively measured with the community modularity Q value of the sample graph in the k-features. An efficient FS method called k-features sample graph feature selection is presented. A key property of this approach is that the discriminative cues of a feature subset with the maximum relevant in-dependency among features can be accurately determined. This community modularity-based method is then verified with the theory of k-means cluster. Compared with other state-of-the-art methods, the proposed approach is more effective, as verified by the results of several experiments.

A Feature Selection Approach for Segmentation of Very High-Resolution Satellite Images

Most of the feature selection (fs) methods in the literature determine features that are appropriate only for a given dataset. In contrast, in this paper a FS method that is not dependent to a specific dataset is proposed. In this regard, the effective feature types based on reasonable facts are predefined and appropriate candidate features for each feature type are selected. In proposed method, the features selected from a single labeled image can be used in segmentation of images captured by different satellites with similar spatial resolution. The selected feature types contain spatial and spectral features. The selected features are applied for segmentation of the images captured by QuickBird and GeoEye satellites and obtained results of proposed method are compared with well-known FS methods. Using different evaluation measures, our comparison shows the efficiency of the proposed method in providing better segmentation compared to other FS methods that are presented in this paper.

Intraoperative frozen section histological analysis of resection samples is useful for the control of primary lesions in patients with oral squamous cell carcinoma.

To ensure reliable surgical margins, intraoperative frozen section histological analysis (FS) has been performed since October, 2005 as follows: i) the orientation at the anatomical position and extent of the tumor are shared between oral pathologists and oral surgeons using imaging evaluations and pathological pictures and the planned site of sampling for intraoperative FS is confirmed; ii) a tumor team is organized and the team marks the tumor area and sets the resection range to correct the setting errors of the resection range among operators; iii) vital Lugol staining is applied to the lesion prior to tumor resection, the surgical margin is set based on the non-stained region and the extent of the tumor is macroscopically confirmed in the maximum cross-sectional surface of the resected specimen; and iv) FS is performed using samples from resected specimens to confirm the mucoepithelium and safety margin of the deep stump. The aim of this study was to evaluate the usefulness of our FS method. The treatment outcomes of oral squamous cell carcinoma were retrospectively investigated in patients treated prior to (Group 1) and after (Group 2) the introduction of our FS method. The recurrence rate of the primary lesions was high (17.3%) in Group 1, but decreased significantly in Group 2 (6.9%). Regarding clinicopathological factors, the condition of the surgical margins was associated with recurrence of the primary lesion in Group 1, but not in Group 2. In conclusion, our FS method appears to be useful for resecting tumors with reliable safety margins.

Exploring the diversity in cluster ensemble generation: Random sampling and random projection

Cluster ensemble first generates a large library of different clustering solutions and then combines them into a more accurate consensus clustering. It is commonly accepted that for cluster ensemble to work well the member partitions should be different from each other, and meanwhile the quality of each partition should remain at an acceptable level. Many different strategies have been used to generate different base partitions for cluster ensemble. Similar to ensemble classification, many studies have been focusing on generating different partitions of the original dataset, i.e., clustering on different subsets (e.g., obtained using random sampling) or clustering in different feature spaces (e.g., obtained using random projection). However, little attention has been paid to the diversity and quality of the partitions generated using these two approaches. In this paper, we propose a novel cluster generation method based on random sampling, which uses the nearest neighbor method to fill the category information of the missing samples (abbreviated as RS-NN). We evaluate its performance in comparison with k-means ensemble, a typical random projection method (Random Feature Subset, abbreviated as FS), and another random sampling method (Random Sampling based on Nearest Centroid, abbreviated as RS-NC). Experimental results indicate that the FS method always generates more diverse partitions while RS-NC method generates high-quality partitions. Our proposed method, RS-NN, generates base partitions with a good balance between the quality and the diversity and achieves significant improvement over alternative methods. Furthermore, to introduce more diversity, we propose a dual random sampling method which combines RS-NN and FS methods. The proposed method can achieve higher diversity with good quality on most datasets.

Testing the mean of an exponential distribution in the presence of outliers

In this study, we propose a simple robust test for the mean of an exponential distribution by using the simplified version of “Forward Search” (FS) method. The FS method is a powerful general method for identifying outliers and their effects on inferences about the hypothesized model. The simulation studies indicate robustness of the testing method and the ability of the procedure to capture the structure of data. Results are presented through the plots which are powerful in revealing the structure of the data.

Combining multiple hypothesis testing and affinity propagation clustering leads to accurate, robust and sample size independent classification on gene expression data

BackgroundA feature selection method in microarray gene expression data should be independent of platform, disease and dataset size. Our hypothesis is that among the statistically significant ranked genes in a gene list, there should be clusters of genes that share similar biological functions related to the investigated disease. Thus, instead of keeping N top ranked genes, it would be more appropriate to define and keep a number of gene cluster exemplars.ResultsWe propose a hybrid FS method (mAP-KL), which combines multiple hypothesis testing and affinity propagation (AP)-clustering algorithm along with the Krzanowski & Lai cluster quality index, to select a small yet informative subset of genes. We applied mAP-KL on real microarray data, as well as on simulated data, and compared its performance against 13 other feature selection approaches. Across a variety of diseases and number of samples, mAP-KL presents competitive classification results, particularly in neuromuscular diseases, where its overall AUC score was 0.91. Furthermore, mAP-KL generates concise yet biologically relevant and informative N-gene expression signatures, which can serve as a valuable tool for diagnostic and prognostic purposes, as well as a source of potential disease biomarkers in a broad range of diseases.ConclusionsmAP-KL is a data-driven and classifier-independent hybrid feature selection method, which applies to any disease classification problem based on microarray data, regardless of the available samples. Combining multiple hypothesis testing and AP leads to subsets of genes, which classify unknown samples from both, small and large patient cohorts with high accuracy.

Discussion: “Factors of Safety for Richardson Extrapolation” (Xing, T., and Stern, F., 2010, ASME J. Fluids Eng., 132, p. 061403)

Discussion: “Factors of Safety for Richardson Extrapolation” (Xing, T., and Stern, F., 2010, ASME J. Fluids Eng., 132, p. 061403)

Conformation of self-assembled porphyrin dimers in liposome vesicles by phase-modulation 2D fluorescence spectroscopy

By applying a phase-modulation fluorescence approach to 2D electronic spectroscopy, we studied the conformation-dependent exciton coupling of a porphyrin dimer embedded in a phospholipid bilayer membrane. Our measurements specify the relative angle and separation between interacting electronic transition dipole moments and thus provide a detailed characterization of dimer conformation. Phase-modulation 2D fluorescence spectroscopy (PM-2D FS) produces 2D spectra with distinct optical features, similar to those obtained using 2D photon-echo spectroscopy. Specifically, we studied magnesium meso tetraphenylporphyrin dimers, which form in the amphiphilic regions of 1,2-distearoyl-sn-glycero-3-phosphocholine liposomes. Comparison between experimental and simulated spectra show that although a wide range of dimer conformations can be inferred by either the linear absorption spectrum or the 2D spectrum alone, consideration of both types of spectra constrain the possible structures to a "T-shaped" geometry. These experiments establish the PM-2D FS method as an effective approach to elucidate chromophore dimer conformation.

659 PHASE 2 TRIAL OF NEOADJUVANT LHRH PLUS ESTRAMUSTINE FOR HIGH RISK PROSTATE CANCER FOLLOWED BY RADICAL PROSTATECTOMY

You have accessJournal of UrologyProstate Cancer: Advanced II1 Apr 2010659 PHASE 2 TRIAL OF NEOADJUVANT LHRH PLUS ESTRAMUSTINE FOR HIGH RISK PROSTATE CANCER FOLLOWED BY RADICAL PROSTATECTOMY Takuya Koie, Teppei Okamoto, Kengo Imanishi, Naoki Sugiyama, Yuichiro Suzuki, Kazuyuki Mori, Shigemasa Kudoh, Takahiro Yoneyama, Yasuhiro Hashimoto, Noritaka Kamimura, and Chikara Ohyama Takuya KoieTakuya Koie More articles by this author , Teppei OkamotoTeppei Okamoto More articles by this author , Kengo ImanishiKengo Imanishi More articles by this author , Naoki SugiyamaNaoki Sugiyama More articles by this author , Yuichiro SuzukiYuichiro Suzuki More articles by this author , Kazuyuki MoriKazuyuki Mori More articles by this author , Shigemasa KudohShigemasa Kudoh More articles by this author , Takahiro YoneyamaTakahiro Yoneyama More articles by this author , Yasuhiro HashimotoYasuhiro Hashimoto More articles by this author , Noritaka KamimuraNoritaka Kamimura More articles by this author , and Chikara OhyamaChikara Ohyama More articles by this author View All Author Informationhttps://doi.org/10.1016/j.juro.2010.02.1046AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail INTRODUCTION AND OBJECTIVES Effective treatment for high risk prostate cancer (PC) has not yet been established. We examined efficacy and safety of neo-adjuvant therapy with LHRH agonist + estramustine (EMP) (LHRH+EMP) followed by retropubic radical prostatectomy (RRP). METHODS Risk stratification was performed according to the D□fAmico□fs method. Eighty seven patients with high risk PC were enrolled onto this trial from September 2005 and August 2009. The LHRH + EMP therapy included administration of LHRH and 280 mg/day EMP for 6 months prior to the RRP. Post-therapy pathological evaluation to no residual disease at RRP (pT0), and PSA-free survival were endpoints of interest. RESULTS The average age of the patients was 67.2 years (range 49 to 78 years), and mean initial PSA level was 22.5 ng/ml (range 4.2 to 95.6 ng/ml). The median Gleason sum score was 8 (range 6 to 9), and clinical stage T2c or greater was 76 patients (87.4%). Eighty-six (98.9%) completed 6 months of LHRH+EMP neo-adjuvant therapy without delay for RP. The pT0 was achieved in 4 patients (4.6%). The negative surgical margin was 73 patients (83.9%). At a mean follow-up period of 22.2 months, the PSA-free survival was 75.1%. The 2-year PSA-free survival was 83.8% with negative surgical margin, and 35.9% with positive surgical margin (P <0.0001) (Figure). CONCLUSIONS Six months of LHRH+EMP neo-adjuvant therapy followed by RP is safe and effective. Although the present study is not randomized, this treatment may improve PSA-free survival in high-risk prostate cancer. A prospective randomized trial is warranted. Hirosaki, Japan© 2010 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 183Issue 4SApril 2010Page: e258 Advertisement Copyright & Permissions© 2010 by American Urological Association Education and Research, Inc.MetricsAuthor Information Takuya Koie More articles by this author Teppei Okamoto More articles by this author Kengo Imanishi More articles by this author Naoki Sugiyama More articles by this author Yuichiro Suzuki More articles by this author Kazuyuki Mori More articles by this author Shigemasa Kudoh More articles by this author Takahiro Yoneyama More articles by this author Yasuhiro Hashimoto More articles by this author Noritaka Kamimura More articles by this author Chikara Ohyama More articles by this author Expand All Advertisement Advertisement PDF downloadLoading ...

Low-Cycle Fatigue Life Prediction by a New Critical-Plane Method

A series of low-cycle fatigue experiments of axial-torsional loading of variable amplitudes were performed on the tubular specimens of 304 stainless steel. Two models of multiaxial low-cycle fatigue life, KBM and FS method, are evaluated based on the fatigue life data of 304 stainless steel. Rainflow cycle counting and the Liner Damage Rule are used to calculate fatigue damage. It was shown that the part prediction results are nonconservative for the two models. The life prediction is done again based on the weight function critical plane method for the two models. The prediction results are better by using the weight function critical plane method than the previous results for KBM model. But the prediction results are improved little for FS model in spite of the weight function critical plane method being used.

A comparison of the indirect estimate of mean arterial pressure calculated by the conventional equation and calculated to compensate for a change in heart rate.

The standard equation used to calculate mean arterial pressure (MAP) assumes that diastole persists for 2/3 and systole for 1/3 of each cardiac cycle. This ratio is altered when heart rate increases, and therefore we investigated the efficacy of predicting MAP during exercise using non-invasive indirect methods. Eight subjects exercised on a cycle ergometer for 3 minute intervals to elicit heart rates between 100-110, 120-130, 140-150, 160-170, and 180-190 beats/min. In the last minute of each 3 min interval an ECG recording was taken and systolic (SP) and diastolic (DP) blood pressure was measured by manual auscultation. MAP was calculated for each heart rate interval by: MAP=DP+1/3(SP-DP) (method A), and MAP= DP + Fs(SP- DP) (method B), where Fs is the fraction of the cardiac cycle comprising systole, measured from the ECG. Fs increased from 0.35+/-0.049 at rest to 0.47+/-0.039 at a heart rate of 180-190 beats/min. MAP measured by method B was consistently greater than MAP calculated by method A at all heart rates greater than resting heart rate (p<0.01). The error incurred when using the standard MAP equation (method A) to derive MAP during exercise (measured as the percentage difference between method A and B) increased linearly with heart rate (r=0.98). The standard MAP equation should not be applied during exercise, as it does not account for the change in the systolic: diastolic period ratio as heart rate increases.

Advantages of the Fourier space RHF band structure approach: Application to polyoxymethylene using a distributed basis set ofs-type Gaussian functions

In this contribution, we outline the Fourier space-restricted Hartree–Fock (FS–RHF) approach to the calculation of the band structure of polyoxymethylene (POM) using a distributed basis set of s-type Gaussian functions (DSGF) to simulate p-type functions. The band structure results are compared to those obtained using minimal STO-3G basis sets, within the conventional RHF direct space (DS) approach, and subminimal floating spherical Gaussian orbital (FSGO) basis sets. While the FSGO basis sets are unable to describe correctly the oxygen lone pairs and their interactions, the DSGF basis set reproduces qualitatively the features of the band structure observed with the minimal basis set. We show that minor differences between the FS and DS results originate in difficulties of lattice summations within the DS approach, illustrating the advantages of the FS method compared to the conventional DS approach. © 1998 John Wiley & Sons, Inc. Int J Quant Chem 70: 1045–1054, 1998

Advantages of the Fourier space RHF band structure approach: Application to polyoxymethylene using a distributed basis set of s‐type Gaussian functions

In this contribution, we outline the Fourier space-restricted Hartree–Fock (FS–RHF) approach to the calculation of the band structure of polyoxymethylene (POM) using a distributed basis set of s-type Gaussian functions (DSGF) to simulate p-type functions. The band structure results are compared to those obtained using minimal STO-3G basis sets, within the conventional RHF direct space (DS) approach, and subminimal floating spherical Gaussian orbital (FSGO) basis sets. While the FSGO basis sets are unable to describe correctly the oxygen lone pairs and their interactions, the DSGF basis set reproduces qualitatively the features of the band structure observed with the minimal basis set. We show that minor differences between the FS and DS results originate in difficulties of lattice summations within the DS approach, illustrating the advantages of the FS method compared to the conventional DS approach. © 1998 John Wiley & Sons, Inc. Int J Quant Chem 70: 1045–1054, 1998