Update README.md

README.md

@@ -3,7 +3,7 @@ Basin Inflow: Regional Hydroclimate Deep Neural Network
 
 Overview
 ----
-Basin Inflow is an implementation of a deep neural network for forecasting reservoir inflow in the American River Basin located in the Sacramento, California region. The proposal model is composed of stacked LSTM layers, with regularization applied to both (L2, dropout, and recurrent dropout), followed by a densely connected layer. Model application and evaluation uses river basin data with temporal coverage from 2008-2022. Across the basin and along forks of the river, daily precipitation, temperature, snow water content and depth, and river discharge and stage are employed to predict local reservoir inflow. Data was pre-processed with a 30-day exponential moving average to give recent weather events more weight in the forecast, normalized, and made stationary for training. 14 years of data were separated into three sets: training set (2008-2017), validation set (2018–2020), and test set (2020-2022). The model achieves a lower MAE score (0.036) in inflow prediction than the baseline model. 
+Basin Inflow is an implementation of a deep neural network for forecasting reservoir inflow in the American River Basin. The proposal model is composed of stacked LSTM layers, with regularization applied to both (L2, dropout, and recurrent dropout), followed by a densely connected layer. Model application and evaluation uses river basin data with temporal coverage from 2008-2022. Across the basin and along forks of the river, daily precipitation, temperature, snow water content and depth, and river discharge and stage are employed to predict local reservoir inflow. Data was pre-processed with a 30-day exponential moving average to give recent weather events more weight in the forecast, normalized, and made stationary for training. 14 years of data were separated into three sets: training set (2008-2017), validation set (2018–2020), and test set (2020-2022). The model achieves a lower MAE score (0.036) in inflow prediction than the baseline model. 
 
 
 Land Acknowledgement