Thermal Interface Material Impact Prediction on LED Heat Dissipation: A Machine Learning-Based Performance Modeling Framework

This dataset contains time-series temperature measurements collected from an E27 LED retrofit lamp under controlled climatic chamber conditions (25°C, 50% relative humidity). The purpose of the dataset is to evaluate the impact of thermal interface material (TIM) application on LED heat dissipation behavior using machine learning models. Data are collected at 30-second intervals and include two experimental configurations:1. Paste (TIM applied)2. No-Paste (TIM not applied) Raw input variables:- IC: Integrated Circuit temperature (°C)- CAP: Capacitor temperature (°C)- COIL: Coil temperature (°C)- LPCB: LED PCB surface temperature (°C)- timestamp: 30-second time index The dataset includes:- Raw CSV files- Preprocessing scripts- Feature engineering modules- Window generation utilities- Scaling parameters- Model training scripts All temperatures are reported in Celsius. The dataset supports reproducible experiments for sequence-based forecasting models such as LSTM and hybrid ML architectures.