Weather & AQI Prediction Model
Forecasting environmental metrics via time-series analysis
completed3 months
#ml
Overview
A machine learning pipeline utilizing both ARIMA and LSTM architectures to predict localized weather variations and Air Quality Index on a custom proprietary dataset.
System Architecture
Raw Dataset
Preprocessing (Pandas/NumPy)
Model Training (ARIMA/LSTM)
Evaluation
TFLite Export
Tech Stack
Py
PythonLanguageTe
TensorFlowMLPa
Pandas/NumPyData ScienceKey Features
- Custom sensor dataset (not public)
- ARIMA time-series model
- LSTM deep learning model
- 13 input parameters → 5 future predictions
- TFLite model export for Android
- Model comparison & accuracy metrics
Project Highlights
LSTM + ARIMA
Models
TFLite
Format
Concept Execution
~/weather-aqi-prediction/execute.sh
❯initializing system instance: Weather & AQI Prediction Model...
❯loading dependencies from core storage...
❯connecting parameters: [ml]
❯system online. executing main loop.
❯▋
