Resource utilization in computer networks is a key area of research. Due to the growing numbers of communication devices and data being processed in the network, today's 5G networks are no exception. Network slicing provides a data-driven, programmable solution by enabling the use of virtual segments over the same underlying physical network, using techniques such as Software Defined Networking and Network Function Virtualization. This paper focuses on predicting the most suitable network slice for network tenants by utilizing the adaptive neuro-fuzzy inference system as a multiclassifier for incoming network slice requests. Each tenant is assigned a suitable slice based on the requesting device's type and characteristics of the required communication channel. The slices considered are massive machine-type communications, enhanced mobile broadband, and ultra-reliable low-latency communications. Our evaluation of the model and simulation results highlights the effectiveness of the Adaptive Neuro-Fuzzy Inference System in selecting the most suitable network slice type, achieving a prediction accuracy of 99.97% using unseen data with a total increase of bandwidth utilization of 12.6% during simulation. Displaying similar or superior performance to existing Deep Learning and Machine Learning approaches.