The building sector is the largest energy-consuming sector in France. In 2021, this consumption accounted for around 40% of total primary energy consumption. In this context, the French tertiary sector decree sets for the building sector (>1000 m²) a reduction in energy consumption of 40% in 2030, 50% in 2040, and 60% in 2050 [1]. In order to achieve these objectives, it is essential to have a better understanding of the thermal behavior of the building. Furthermore, the existing simulation tools consider the occupants as static participants, whereas their interaction with the building has a significant influence. These factors actually have a high influence on building energy consumption, particularly in large occupancy buildings like educational buildings [2]. To achieve high performance and low energy consumption in the buildings, we propose a study to develop a simple yet robust model to simulate and predict the thermal condition of the classroom using advanced ensemble machine-learning techniques. The paper also proposes the study on optimal placement and number of sensors required for measuring the indoor temperature in the room.