scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

Hybrid prediction model based on BP neural network for lung cancer

01 Dec 2008-pp 532-535
TL;DR: Two hybrid prediction models based on BP neural network, ES (exponential smoothing) and FCM (Fuzzy C-Means) clustering are proposed to predict the possible rate and ages of smokers suffering the lung cancer.
Abstract: Recent researches show that lung cancer owns actual dose-response relationship with calendar-year smoking environment exposure matrix and individual medical record. In this paper, two hybrid prediction models based on BP neural network, ES (exponential smoothing) and FCM (Fuzzy C-Means) clustering are proposed to predict the possible rate and ages of smokers suffering the lung cancer. The BP-ES (Exponential Smoothing) model can exert the superiorities of the time series datum of smoking crowds and other pathogenic factors; and the BPFCM clustering model can reduce the parameter amount and complexity of BP netpsilas training greatly. The experiments show that the accuracy of the hybrid models are enhanced greatly contrasted with single BP neural network, and can work as effective methods for the statistic, analysis and prediction to lung cancer.
Citations
More filters
Journal ArticleDOI
TL;DR: A survey of recent work on computational intelligence approaches that have been applied to prostate cancer predictive modeling, and considers the challenges which need to be addressed.
Abstract: Focus is on computational intelligence methods in prostate cancer predictive modeling.We survey metaheuristic optimisation methods.We review machine learning methods.We consider cancer data of different modalities.We discuss recent advances, challenges and provide future directions. Predictive modeling in medicine involves the development of computational models which are capable of analysing large amounts of data in order to predict healthcare outcomes for individual patients. Computational intelligence approaches are suitable when the data to be modelled are too complex for conventional statistical techniques to process quickly and efficiently. These advanced approaches are based on mathematical models that have been especially developed for dealing with the uncertainty and imprecision which is typically found in clinical and biological datasets. This paper provides a survey of recent work on computational intelligence approaches that have been applied to prostate cancer predictive modeling, and considers the challenges which need to be addressed. In particular, the paper considers a broad definition of computational intelligence which includes metaheuristic optimisation algorithms (also known as nature inspired algorithms), Artificial Neural Networks, Deep Learning, Fuzzy based approaches, and hybrids of these, as well as Bayesian based approaches, and Markov models. Metaheuristic optimisation approaches, such as the Ant Colony Optimisation, Particle Swarm Optimisation, and Artificial Immune Network have been utilised for optimising the performance of prostate cancer predictive models, and the suitability of these approaches are discussed.

73 citations

Journal Article
TL;DR: Compared with the FCNN algorithm, RFCNN is much more robust to outliers in the datasets, and its update rules are derived by using Lagrange optimization theory.
Abstract: In this paper a new robust fuzzy clustering neural networks (RFCNN) is presented to resolve the sensitivity of the fuzzy clustering neural network (FCNN) to outliers in real datasets. The new objective function of RFCNN is obtained by introducing Vapnik’s e-insensitive loss function, and RFCNN’s update rules are derived by using Lagrange optimization theory. Compared with the FCNN algorithm, RFCNN is much more robust to outliers in the datasets. Experimental results demonstrate the effectiveness of RFCNN.

5 citations

Journal ArticleDOI
TL;DR: The random forest algorithm is the most suitable prediction model for predicting medical costs and patient survival with the quantity of data currently available and could provide guidance and help medical institutions improve the efficiency and quality of home medical services for patients with advanced cancer.
Abstract: Background As the number of patients with cancer rises, home care for patients with advanced disease is becoming increasingly important. To provide guidance for home medical services and hospice care, we investigated the basic information and medical service information of patients with advanced cancer receiving home care by using a data mining algorithm to predict the patients’ survival and medical expenses. Methods Data from patients with advanced cancer who received home care in Chongming District (Shanghai, China) between 2016 and 2018 were collected. The medical expenses and survival time of the patients were classified and predicted through the use of random forest algorithms, support-vector machine algorithms, and back-propagation (BP) neural network algorithms. Results The performances of the 3 algorithms in classifying patient survival and predicting medical expenses were compared. The random forest algorithm, support vector machine, and BP neural network in the classification of patient survival had accuracy of 81.94%±6.12%, 74.61%±7.01%, and 72.90%±8.08%, respectively. The standard mean square errors of the regression model for predicting medical expenses were 0.4194±0.2393, 1.1222±0.0648, and 1.2986±0.1762, respectively. Conclusions The random forest algorithm is the most suitable prediction model for predicting medical costs and patient survival with the quantity of data currently available. Further optimization of the random forest algorithm could provide guidance and help medical institutions improve the efficiency and quality of home medical services for patients with advanced cancer.

2 citations


Cites methods from "Hybrid prediction model based on BP..."

  • ...The support vector machine (SVM) (11) and BP neural network (12) algorithms were used to classify and predict the patient’s home medical service duration (days) and to perform a regression analysis of drug costs (Chinese Yuan), respectively (13-15)....

    [...]

References
More filters
Journal ArticleDOI
TL;DR: Differences in the risk of silicosis among cohorts of silica dust-exposed Chinese tin miners, tungsten miners, and pottery workers suggest that silica Dust characteristics, in addition to cumulative respirable silicaDust exposure, may affect the riskof silicotic disease.
Abstract: Background Epidemiological evaluations of the risk of silicosis in relation to exposure to crystalline silica have raised the question of whether different types of silica dust exposures vary with respect to their ability to cause silicosis. The aim of this study is to compare the risk of silicosis among cohorts of silica dust-exposed Chinese tin miners, tungsten miners, and pottery workers and to assess whether gravimetric measurements of respirable silica dust sufficiently determine the risk of silicosis or whether other factors of exposure may play a significant role. Methods Cohorts were selected from 20 Chinese mines and potteries. Inclusion criteria were starting employment after January 1, 1950 and being employed for at least 1 year during 1960–1974 in one of the selected workplaces. Radiological follow-up for silicosis onset was from January 1, 1950 through December 31, 1994. Silicosis was assessed according to the Chinese radiological criteria for diagnosis of pneumoconiosis (as suspect, Stage I, II, or III). Exposure–response relationships were estimated for silicosis of Stage I or higher. Silica dust exposure was estimated in terms of cumulative total dust exposure, calculated from a workplace, job title, and calendar year exposure matrix, and individual occupational histories. Cumulative total dust exposure was converted in two steps into cumulative respirable dust exposure and cumulative respirable silica dust exposure using conversion factors estimated from side-by-side measurements conducted in 1988–89. Results The male cohorts included 4,028 tin miners, 14,427 tungsten miners, and 4,547 pottery workers who had similar onset of employment and duration of follow-up. For a given exposure level, the risk of silicosis was higher for the tin and tungsten than the pottery workers. Conclusion The observed differences in the risk of silicosis among the three cohorts suggest that silica dust characteristics, in addition to cumulative respirable silica dust exposure, may affect the risk of silicosis. Am. J. Ind. Med. 48:1–9, 2005. Published 2005 Wiley-Liss, Inc.

60 citations

Journal ArticleDOI
TL;DR: Results reinforce the recommendation that silica exposure should be halted at an early stage whenever X-ray is suggestive of the disease, and compare workers who continued to be exposed to silica with those who stopped silica Exposure after having received their diagnosis.
Abstract: Background There is a paucity of studies analyzing the effect of continued silica exposure after the onset of silicosis with regard to disease progression. The present study investigates differences in clinical and radiological presentation of silicosis among former workers with a diagnosis of silicosis, and compares workers who continued to be exposed to silica with those who stopped silica exposure after having received their diagnosis. Methods A sample of 83 former gold miners with a median of 21 years from the first diagnoses of silicosis, had their clinical and occupational histories taken and underwent both chest radiography (International Labor Organization standards) and spirometry. Their silica exposure was assessed and an exposure index was created. The main outcome was the radiological severity of silicosis and tuberculosis (TB). The statistical analysis was done by multiple logistic regression. Results Among the 83 miners, 44 had continued exposed to silica after being diagnosed with silicosis. Continuation of silica exposure was associated with advanced radiological images of silicosis (X-ray classification in category 3, OR = 6.42, 95% CI = 1.20–34.27), presence of coalescence and/or large opacities (OR = 3.85, CI = 1.07–13.93), and TB (OR = 4.61, 95% CI = 1.14–18.71). Conclusions Differential survival is unlikely to explain observed differences in silicosis progression. Results reinforce the recommendation that silica exposure should be halted at an early stage whenever X-ray is suggestive of the disease. Am. J. Ind. Med. 49:811–818, 2006. © 2006 Wiley-Liss, Inc.

21 citations


"Hybrid prediction model based on BP..." refers methods in this paper

  • ...If we use single BP neural network to make prediction, the error will be piled up following the prediction steps [2]....

    [...]

Book ChapterDOI
06 Jun 2005
TL;DR: The modified method allows finding the appropriate number of classes and uses improvements introducing in conventional fuzzy c-means algorithm increasing its robustness to the influence of outliers.
Abstract: Proposed method of clustering is based on modified fuzzy c-means algorithm. In the paper features of input data are considered as linguistic variables. Any feature is described by set of fuzzy numbers. Thus, any input data representing a feature is a fuzzy number. The modified method allows finding the appropriate number of classes. Moreover, it uses improvements introducing in conventional fuzzy c-means algorithm increasing its robustness to the influence of outliers.

15 citations


"Hybrid prediction model based on BP..." refers background or methods in this paper

  • ...The artificial neural networks offer advantages such as learning, adoption, fault-tolerance, parallelism, and generalization [7]....

    [...]

  • ...So that the exceptional datum can be removed and the efficiency and accuracy of the FCM-BP hybrid models is enhanced greatly contrasted with single BP neural network [7]....

    [...]

01 Jan 1979

7 citations

Journal Article
TL;DR: Compared with the FCNN algorithm, RFCNN is much more robust to outliers in the datasets, and its update rules are derived by using Lagrange optimization theory.
Abstract: In this paper a new robust fuzzy clustering neural networks (RFCNN) is presented to resolve the sensitivity of the fuzzy clustering neural network (FCNN) to outliers in real datasets. The new objective function of RFCNN is obtained by introducing Vapnik’s e-insensitive loss function, and RFCNN’s update rules are derived by using Lagrange optimization theory. Compared with the FCNN algorithm, RFCNN is much more robust to outliers in the datasets. Experimental results demonstrate the effectiveness of RFCNN.

5 citations


"Hybrid prediction model based on BP..." refers background or methods in this paper

  • ...The artificial neural networks offer advantages such as learning, adoption, fault-tolerance, parallelism, and generalization [ 7 ]....

    [...]

  • ...So that the exceptional datum can be removed and the efficiency and accuracy of the FCM-BP hybrid models is enhanced greatly contrasted with single BP neural network [ 7 ]....

    [...]