With the continuous development and relentless growth of sensor devices across various domains of daily life, the volume of information generated by wireless sensor network systems has steadily increased. This includes anomalous data, which enhances the effectiveness of environmental monitoring and control. This paper proposes an integrated framework comprising multiple components, ranging from data collection, storage, and preprocessing to the application of machine learning and deep learning methodologies for training models on the acquired dataset. Additionally, the framework employs advanced techniques such as Neural Architecture Search and Grid Search for model optimization and hyperparameter tuning. Moreover, the authors utilize a hybrid approach to forecast an adequate volume of information within a near-future timeframe.