PROGNOSIS OF HYPER TRIGLYCERIDES USING DATA SCIENCE AND MACHINE LEARNING

Abstract

The main aim is to propose a model for the prediction of hyper triglycerides using supervised machine learning algorithms based on several risk factors, such as heart illnesses, particularly CAD, stroke, liver, renal, and other chronic diseases. The model's significant feature is that it enables medical professionals to reassess the related risk and provide appropriate guidelines and medications to manage or for prevention of its occurrence. According to the performance research, data preparation is vital in order to produce a model that is both accurate and efficient. Thus, the results of the test proved that logistic regression was efficient, with recall of 0.98 and F1-measure of 0.99 with 99% accuracy. The future model can be evolved through the deep learning techniques to predict triglyceride levels even in infants and newborns. This will enable the development of more precise prognoses and diagnoses with regard to remedies by analyzing large datasets of patient information including genomic data and other medical records.

Specification

Description:FIELD OF INVENTION
Early Triglycerides are a type of fat. They are the most common type of fat in our body. They emanate from foods, especially butter, oils, and other fats we eat and also come from extra calories. These are the calories that we eat, but our body does not need right away. Our body naturalizes these extra calories into triglycerides and stores them in fat cells. When our body needs energy, it disseminates the triglycerides. Our VLDL cholesterol particles carry the triglycerides to our tissues. Hyper triglycerides can increase the risk of heart diseases in particular, CAD, stroke, liver, kidney and other chronic diseases. In recent years Data science is one of the progressing demesne due to the profusion of data sources and resulting data. The realm of healthcare is substantially ameliorated from Data science and Machine Learning applications because of these intuitive solutions. Using Data science techniques and Machine learning algorithms with ANN, we can prognoses the disease. The WHF dossier says that every year nearly 4.4 million death occurs due to heart diseases and WHO says that nearly 2.6 million deaths occur due to cholesterol.
BACKGROUND OF THE INVENTION
Goran Walldiusa and Ingmar Junger (2007) demonstrated to compare the potential of high-density lipoprotein (HDL) cholesterol and apolipoprotein (apo) A-I, the major protein in HDL particles, in predicting cardiovascular risk. Pros and cons for using these risk markers are discussed. Jing Ma et al., (2017) analyzed the relationship of triglyceride (TG) and cholesterol (TC) with indexes of liver function and kidney function, and to develop a prediction model of TG, TC in overweight people. Sajida Perveen et al., (2018) presented this research is to develop machine learning based method in order to identify individuals at an increased risk of developing Non-Alcoholic Fatty Liver Disease using risk factors of Metabolic Syndrome validate the relative ability of quantitative score. Nahuel García-D'urso et al., (2022) described the machine learning approach to predict cholesterol levels using non-invasive and easy-to-collect data is presented. Specially, it uses clinical and anthropometric data gathered by nutritionists during weight loss intervention (dieting) periods.
SUMMARY OF THE INVENTION
In this invention, we propose a model for the prediction of hyper triglycerides using supervised machine learning algorithms based on several risk factors, such
Fig 1.1 Architecture of the Invention
1.1. Raw Data Sets:
Data that have been gathered from sources like Google form, kaggle and other health related data sets from different sources are known as raw data or primary data. To forecast the HTG and HC, this system's data was collected from a variety of sources to compile the Dataset from many sources. The first phase in the Data Science process is data collection for the ML model's training.
1.2. Data Preprocessing:
The classifier purges the data set of unnecessary information. Thus, a number of preprocessing stages are used to assure data validity, and data cleaning techniques are used to ensure data quality. Some instances of data cleaning techniques involve getting rid of redundant data, preventing typos, handling missing values, data imputation, etc. We decided to eliminate outliers (i.e., occurrences with missing and invalid feature values) from the present data set.
1.3. Feature selection:
In machine learning, the process of feature selection identifies critical elements in a dataset to enhance the model's performance and interpretability. It is mainly for characteristic's significance to the target variable and to compute scores for each feature[12]. Based on their ratings, it chooses a subset of the most significant features and uses them to train the predictive model.
1.4. New Data Set:
The new data is relevant, reliable, and consistent with the existing data. All outliers, duplicates, and redundant data have been removed from the new data set. And thus the required or necessary variables like
1.5. Machine Learning Algorithm:
In this system, the model is trained using Supervised Machine learning algorithms[1]. Supervised learning is a method of machine learning in which the output is predicted by the machines using well-labelled training data that has been used to train the machines. The term "labelled data" refers to input data that has already been assigned the appropriate output.
In supervised learning, the training data that is given to the computers serves as the supervisor, instructing them on how to correctly predict the output. It employs the same approach that a pupil would learn under a teacher's guidance.
The method of supervised learning involves giving the machine learning model appropriate input data as well as the output data. Identifying a mapping function to link the input variable (x) with the output variable (y) is the goal of a supervised learning algorithm. It also encompasses applications in the real world such as risk assessment, image categorization, fraud detection, spam filtering, etc.
1.6 Random Forest Algorithm:
The supervised learning method includes the well-known machine learning algorithm Random Forest. It can be applied to ML Classification and Regression issues[2]. Its foundation is the idea of ensemble learning, which is the process of mixing various classifiers to solve a challenging problem and enhance the performance of the model.
Random Forest is a classifier that uses numerous decision trees based on different subsets of the provided dataset and aggregates the results to increase the dataset's predicted accuracy. Rather than depending on a single decision tree, the random forest uses forecasts from each tree and predicts the result based on the votes of the majority of predictions. Larger the number of trees in the forest, higher the accuracy and overfitting are hence prevented.
Fig 1.2 Random Forest Algorithm process
The Figure 1.2 Considering the fact that the random forest aggregates several trees to estimate the dataset's class, it is possible that some decision trees will predict the correct result while others won't. But when evaluated together, each tree accurately predicts the outcome.
The OOB score is a measure of the model's performance on unseen data, which means that the data are not used in the training of the model and are used to provide an unbiased estimate of the model's performance. The OOB score is calculated for ensemble models and the error rate varies from 0.036 to 0.042 and above. The OOB error rate for 70 trees is 0.03774 with score of 0.964033.
Figure 1.3: OOB Error rate with the trees
Classification report of Random Forest Algorithm , Claims:The system provides the prediction of hyper triglycerides using supervised machine learning algorithms based on several risk factors, such as heart illnesses, particularly CAD, stroke, liver, renal, and other chronic diseases.
2. The system's significant feature is that it enables medical professionals to reassess the related risk and provide appropriate guidelines and medications to manage or for prevention of its occurrence.
3. According to the performance research, data preparation is vital in order to produce a model that is both accurate and efficient.
4. The system aggregates patient data for comprehensive analysis, providing healthcare
professionals with valuable insights into patient trends, treatment efficacy and potential
complications which can improve decision-making.

Documents