The study brings an effective convolutional neural network model for classification of clinical data which is highly imbalanced. The data is obtained from NHANES with the aim to predict Coronary Heart Disease (CHD). Using two step approach i.e. performing LASSO Regression for feature selection and then applying CNN on selected data, classification is performed to achieve better prediction. Most of the machine learning models can become vulnerable when introduced to imbalanced data i.e. huge difference in proportion of class type. This simple two layered CNN performs better even in the presence of imbalance in dataset. The two step approach is in a way that first it use Least Absolute Shrinkage and Selection Operator (LASSO) regression method for the purpose of feature selection and then use CNN to predict the CHD. The proposed CNN model classify status of CHD with the accuracy of 85.70%.
Input variables : Age, systolic and diastolic blood pressure, BMI, Weight, White blood cells, Hemoglobin, Platelet count, Albumin, Diabetes status etc.
Output Variables : Heart Disease Status
| Statistical | : | Somers D | Accuracy | Precision and Recall | Confusion Matrix | F1 Score | Roc and Auc | Prevalence | Detection Rate | Balanced Accuracy | Cohen's Kappa | Concordance | Gini Coefficent | KS Statistic | Youden's J Index |
| Infrastructure | : | Log Bytes | Logging/User/IAMPolicy | Logging/User/VPN | CPU Utilization | Memory Usage | Error Count | Prediction Count | Prediction Latencies | Private Endpoint Prediction Latencies | Private Endpoint Response Count |
Visit Model : github.com
Additional links : arxiv.org
| Model Category | : | Public |
| Date Published | : | April, 2020 |
| Healthcare Domain | : | Provider |
| Code | : | github.com |
| Health Risk Management |
| Chronic Care Management |