In cold storage facilities, monitoring the inside temperature is necessary to preserve the products. The larger the ambient is, the more sensors are needed to cover all the space. In a cenario where 2 sensors are installed, could we predict one only by measuring the other since they are highly correlated? This notebook aims to study the possibility to predict such temperature. We will use a LSTM neural network due to its capacity of learning temporal behaviours on a time series.

Suggestions and feedbacks are welcome.