A comparison between logistic regression and neural networks in a constructed response item study

SummaryObjective: To describe and compare logistic regression and neural network modeling strategies to predict hospital admission or transfer following initial presentation to Emergency Department (ED) triage with and without the addition of natural language processing elements.Methods: Using data from the National Hospital Ambulatory Medical Care Survey (NHAMCS), a cross-sectional probability sample of United States EDs from 2012 and 2013 survey years, we developed several predictive models with the outcome being admission to the hospital or transfer vs. discharge home. We included patient characteristics immediately available after the patient has presented to the ED and undergone a triage process. We used this information to construct logistic regression (LR) and multilayer neural network models (MLNN) which included natural language processing (NLP) and principal component analysis from the patient’s reason for visit. Ten-fold cross validation was used to test the predictive capacity of each model and receiver operating curves (AUC) were then calculated for each model.Results: Of the 47,200 ED visits from 642 hospitals, 6,335 (13.42%) resulted in hospital admission (or transfer). A total of 48 principal components were extracted by NLP from the reason for visit fields, which explained 75% of the overall variance for hospitalization. In the model including only structured variables, the AUC was 0.824 (95% CI 0.818-0.830) for logistic regression and 0.823 (95% CI 0.817-0.829) for MLNN. Models including only free-text information generated AUC of 0.742 (95% CI 0.7310.753) for logistic regression and 0.753 (95% CI 0.742-0.764) for MLNN. When both structured variables and free text variables were included, the AUC reached 0.846 (95% CI 0.839-0.853) for logistic regression and 0.844 (95% CI 0.836-0.852) for MLNN.Conclusions: The predictive accuracy of hospital admission or transfer for patients who presented to ED triage overall was good, and was improved with the inclusion of free text data from a patient’s reason for visit regardless of modeling approach. Natural language processing and neural networks that incorporate patient-reported outcome free text may increase predictive accuracy for hospital admission.

Download Full-text

An Optimal Model of Financial Distress Prediction: A Comparative Study between Neural Networks and Logistic Regression

Risks ◽

10.3390/risks9110200 ◽

2021 ◽

Vol 9 (11) ◽

pp. 200

Author(s):

Youssef Zizi ◽

Amine Jamali-Alaoui ◽

Badreddine El Goumi ◽

Mohamed Oudgou ◽

Abdeslam El Moudden

Keyword(s):

Neural Networks ◽

Logistic Regression ◽

Financial Distress ◽

Prediction Models ◽

Type I ◽

Return On Assets ◽

Logistic Regression Models ◽

The Face ◽

One Year ◽

Type Ii Errors

In the face of rising defaults and limited studies on the prediction of financial distress in Morocco, this article aims to determine the most relevant predictors of financial distress and identify its optimal prediction models in a normal Moroccan economic context over two years. To achieve these objectives, logistic regression and neural networks are used based on financial ratios selected by lasso and stepwise techniques. Our empirical results highlight the significant role of predictors, namely interest to sales and return on assets in predicting financial distress. The results show that logistic regression models obtained by stepwise selection outperform the other models with an overall accuracy of 93.33% two years before financial distress and 95.00% one year prior to financial distress. Results also show that our models classify distressed SMEs better than healthy SMEs with type I errors lower than type II errors.

Download Full-text

Performance Comparison of Internal Accounting Control Assessment Models Applying Logistic Regression and Neural Networks

Korea International Accounting Review ◽

10.21073/kiar.2012..46.001 ◽

2012 ◽

Vol null (46) ◽

pp. 1-30

Author(s):

김명종

Keyword(s):

Neural Networks ◽

Logistic Regression ◽

Performance Comparison ◽

Internal Accounting Control ◽

Accounting Control ◽

Internal Accounting

Download Full-text

Identification of Long Non-coding RNA Isolated From Naturally Infected Macrophages and Associated With Bovine Johne's Disease in Canadian Holstein Using a Combination of Neural Networks and Logistic Regression

Frontiers in Veterinary Science ◽

10.3389/fvets.2021.639053 ◽

2021 ◽

Vol 8 ◽

Author(s):

Andrew Marete ◽

Olivier Ariel ◽

Eveline Ibeagha-Awemu ◽

Nathalie Bissonnette

Keyword(s):

Neural Networks ◽

Logistic Regression ◽

Differential Expression Analysis ◽

Johne's Disease ◽

Classification Systems ◽

Johne’S Disease ◽

Potential Candidate ◽

Rna Seq ◽

Non Coding Rna ◽

Long Non Coding Rna

Mycobacterium avium ssp. paratuberculosis (MAP) causes chronic enteritis in most ruminants. The pathogen MAP causes Johne's disease (JD), a chronic, incurable, wasting disease. Weight loss, diarrhea, and a gradual drop in milk production characterize the disease's clinical phase, culminating in death. Several studies have characterized long non-coding RNA (lncRNA) in bovine tissues, and a previous study characterizes (lncRNA) in macrophages infected with MAP in vitro. In this study, we aim to characterize the lncRNA in macrophages from cows naturally infected with MAP. From 15 herds, feces and blood samples were collected for each cow older than 24 months, twice yearly over 3–5 years. Paired samples were analyzed by fecal PCR and blood ELISA. We used RNA-seq data to study lncRNA in macrophages from 33 JD(+) and 33 JD(–) dairy cows. We performed RNA-seq analysis using the “new Tuxedo” suite. We characterized lncRNA using logistic regression and multilayered neural networks and used DESeq2 for differential expression analysis and Panther and Reactome classification systems for gene ontology (GO) analysis. The study identified 13,301 lncRNA, 605 of which were novel lncRNA. We found seven genes close to differentially expressed lncRNA, including CCDC174, ERI1, FZD1, TWSG1, ZBTB38, ZNF814, and ZSCAN4. None of the genes associated with susceptibility to JD have been cited in the literature. LncRNA target genes were significantly enriched for biological process GO terms involved in immunity and nucleic acid regulation. These include the MyD88 pathway (TLR5), GO:0043312 (neutrophil degranulation), GO:0002446 (neutrophil-mediated immunity), and GO:0042119 (neutrophil activation). These results identified lncRNA with potential roles in host immunity and potential candidate genes and pathways through which lncRNA might function in response to MAP infection.

Download Full-text

Comparison of Neural Networks and Logistic Regression in Assessing the Occurrence of Failures in Steel Structures of Transmission Lines

The Open Electrical & Electronic Engineering Journal ◽

10.2174/1874129001610010011 ◽

2016 ◽

Vol 10 (1) ◽

pp. 11-26

Author(s):

A.C.G Bissacot ◽

S.A.B Salgado ◽

P.P Balestrassi ◽

A.P Paiva ◽

A.C Zambroni Souza ◽

...

Keyword(s):

Neural Networks ◽

Logistic Regression ◽

Transmission Lines ◽

Steel Structures ◽

Roc Curves ◽

Operating Voltage ◽

Operating Characteristics ◽

Metal Structures ◽

Land Type ◽

Classic Technique

In this work, we evaluate the probability of falling metal structures from transmission lines. It is our objective to extract knowledge about which variables influence the mechanical behavior of the operating lines and can be used to diagnose potential falling towers. Those pieces of information can become a basis for directing the investments of reinforcement structures, avoiding the occurrence of long turn offs and high costs as a consequence of damage to towers of transmission lines. The results are obtained using the history of 181 metal structures currently in operation in the state of Paraná/Brazil. For the classification of transmission lines susceptible to failures it is proposed to identify the most likely lines considering the following parameters: operating voltage, wind and relief of the region, air masses, temperature, land type, mechanical capacity, function and foundation structure. The classic technique of classifying binary events used in this type of problem is the logistic regression (LR). The more recent technique for classification, using Artificial Neural Networks (ANN) can also be applied. The results are compared through the area under receiver operating characteristics (ROC) curves.

Download Full-text