Explainable Data Driven Digital Twins for Predicting Battery States in Electric Vehicles

Project Code :TCMAPY1378

Objective

This work aims to develop an advanced Digital Twin model for predicting battery states (SOC and SOH) in electric vehicles, using algorithms like DNN, LSTM, CNN, SVR, Random Forests, and XGBoost. The goal is to enhance model accuracy, interpretability through explainable AI, and compare its performance with conventional methods, providing a smarter, more transparent solution for battery management and durability assessment in the EV industry.

Abstract

Abstract

The main goal of this work is to predict medication toxicity with the help of deep learning models and consider opioids prescriptions only. As the dataset used has 25k records and 256 features, Recurrent Neural Networks (RNN), Random Forest, XGBoost, Voting Classifier increase the chances of prediction. To enhance the feature space in the dataset, various features regarding medications that are frequently prescribed include ACETAMINOPHEN, GABAPENTIN, and LEVOTHYROXINE among others; the target feature is known as the ‘Opioid Prescriber’. With the goal of detecting whether a prescriber is subscribing opioids and to minimize the threats related tole opioid toxicity, this work tries to ascertain associative structures among distinct medications. The integration of machine learning and deep learning will form a strong foundation, so further enhancing prescription safety and eliminating opioid toxicities.

Keywords: Medication Toxicity Forecasting: Deep Learning for the Prediction of Opioid Prescriber Random Forest, eXtreme Gradient Boosting, Voting Classifier Prescription Pattern Analysis: Toxicity Levels of Medication Ensemble Learning

NOTE: Without the concern of our team, please don't submit to the college. This Abstract varies based on student requirements.