Forecasting hourly PM_2.5 concentration with an optimized LSTM model

Keyphrases

• Particulate Matter 2.5 (PM2.5) 100%
• PM2.5 Concentration 100%
• Long-Short 100%
• Memory Models 100%
• Root Mean Square Error 66%
• Air Quality 66%
• Air Quality Monitoring 66%
• Taiwan 33%
• Urban Areas 33%
• Seasonal Variation 33%
• Public Health 33%
• Machine Learning 33%
• Strong Correlation 33%
• Deep Learning Methods 33%
• Result Prediction 33%
• Aerosol Optical Depth 33%
• Deep Learning Algorithm 33%
• Pearson Correlation Coefficient 33%
• Timely Information 33%
• Model Setting 33%
• Spatial Representation 33%
• Urban Air 33%
• PM2.5 Prediction 33%
• Polluted Area 33%
• Sensitivity Assessment 33%
• Highest Root 33%
• Predictable Information 33%
• Model Predictability 33%
• Training Units 33%

Engineering

• Long Short-Term Memory 100%
• Particular Matter 2.5 100%
• Air Quality 66%
• Root Mean Square Error 33%
• Deep Learning 33%
• Input Parameter 16%
• Powerful Tool 16%
• Model Input 16%
• Aerosol Optical Depth 16%
• Quality Assessment 16%
• Pearsons Linear Correlation Coefficient 16%

Earth and Planetary Sciences

• Particular Matter 2.5 100%
• Air Quality 50%
• Root-Mean-Square Error 33%
• Taiwan 16%
• Annual Variation 16%
• Correlation Coefficient 16%
• Machine Learning 16%
• Air Quality Monitoring 16%