A wireless sensor network (WSN) consists of spatially distributed independent devices to monitor physical or environmental conditions. Within the communication range of WSN, the sensor nodes transmit the sensed data to the base station. During the data collection, some amount of energy is dissipated. So, an energy-efficient data collection scheme is required for extending the network lifetime. A novel Energy-Aware Best First indexive Weighted Random Decision Forest Classification (EABFIWRDFC) technique is introduced for efficient data collection in WSN with higher accuracy and lesser time consumption. Initially in the ABFIWRDFC technique, a number of sensor nodes and threshold energy levels are initialized for performing the data collection. After that, the energy level of the sensor node is calculated. Camargo's indexive Instant Weighted Random Decision Forest algorithm is introduced to classify the higher energy nodes with majority votes for performing data collection. Random Decision Forest Classifier is an ensemble technique that uses the Camargo's indexive best first decision trees as weak learner to distinguish the higher energy and lesser energy sensor nodes. Then the lesser energy nodes transmit the sensed data packets to the nearest higher energy nodes by using Manhattan distance. After that, the sink node collects the data from the higher energy sensor node with higher accuracy and lesser time consumption. Experimental evaluation is carried out on factors such as energy consumption, packet delivery ratio, packet loss rate, throughput, and delay. The observed simulation results illustrate that the EABFIWRDFC technique efficiently improves the data delivery, throughput and minimizes the energy consumption, loss rate as well as delay than the conventional methods.