Open

Hi, I'm Shiva Dhanush, and I attend RV University to study computer science. After completing my first year, I can't wait to learn more about artificial intelligence and software development. I worked on a number of personal projects throughout my first year, including a basic Python chatbot and a productivity smartphone app. In addition, I enjoy learning new skills and pushing myself in hackathons and coding contests.

Poster Abstract

Acute liver failure (ALF) is a critical medical condition requiring prompt diagnosis and intervention. This project aims to develop a predictive model for ALF using machine learning techniques, specifically HPCC Systems' Generalized Linear Model (GLM). The dataset used for this analysis is available Here.. This includes various clinical and demographic attributes such as bilirubin levels, age, and prothrombin time, which are instrumental in identifying ALF cases.

The primary objective of this project is to enhance the early detection of ALF by building a robust classification model that can accurately predict the onset of this disease based on patient data. The dataset encompasses diverse features that contribute to the pathogenesis of liver failure, enabling a comprehensive analysis of the factors leading to ALF.

The focus is twofold: firstly, to visualize the significant clinical features associated with ALF, providing insights into their contributions and interactions; and secondly, to construct a predictive model that evaluates new patient data to forecast ALF risk. This model aims to support healthcare providers by offering a tool for early diagnosis, thereby improving patient outcomes through timely medical intervention.

HPCC Systems’ capabilities in managing and visualizing complex datasets will be crucial in conveying the findings of this analysis and in building a reliable predictive framework for acute liver failure.

This is a work in progress and when tested locally it has given an accuracy of 0.91. However efforts ongoing to spray it on the HPCC cluster and train the model using HPCC ml bundles.

Realtime results can be given for individual patients data.

Future enhancements would be to build a website where liver problems could be diagnosed.

Presentation

In this Video Recording, Shiva provides a tour and explanation of his poster content.

Predicting Acute Liver Failure using Machine Learning:

Click on the poster for a larger image.

Open