Cellulose is considered to be one of the most underutilized biomass available on earth. These cellulosic resources, if utilized as the precursor of food, feed and biofuel can meet up the ever-increasing demands for food and energy. Since cellulase is an enzyme complex, fermentation of cellulose from agro-wastes and industrial effluents is a complex event. The physicochemical parameters are generally optimized in "one at a time" mode in fed-batch culture. But with the advent of time, statistical and mathematical modelling is used, like response surface methodology (RSM), artificial neural network (ANN), machine learning algorithm, and genetic algorithm to improve enzyme production. In bioreactor-based cellulase production, a LabVIEW-based intelligent system for monitoring bio-processing is used for the optimization of the target parameters. RSM and ANN are high-quality prediction mathematical models but ANN shows its superiority in context to the fitting of data as well as its estimation capabilities. The difference of ANN concerning RSM is its requirement of a large number of trained data. This review provides a comprehensive study of literature in context to various advanced mechanisms for optimization of cellulase production.