Abstract：A Machine Learning Based Book Availability Prediction Model for Library Management System The Library Management System supports numerous users every day and yet many users cannot avail books in real time for their useFor a userit will be very beneficial to have a system which can predict the possible availability of the issued booksIn this paperMachine learning is used on the data obtained from the library to predict the date for book availabilityRandom forestsupport vector and neural network are used and the result trend are compared using keres and SKlearnFrom the studythe result shows that it is possible to know and govern the availability of the books issuedThe learned model can then be used to predict the availability of the bookHoweverthe analysis accuracy is reduced when the quality of library data is incompleteIn this studystreamline machine learning algorithms for effective prediction of books in library system is usedThe experiment of the modified prediction models over real-life library data collected from Central library of Central Institute of Technology KokrajharCLCITKwas usedTo overcome the difficulty of incomplete dataa latent factor model to reconstruct the missing data was usedThis study is a proposal for a new model using different machine learning method and to compare performance among them and to identify the more suitable method for the prediction system of book availability in librariesLibrary AutomationFocus on predicting the availability of booksThis study uses machine learning to analyze the library data to predict the availability of booksUse random forestssupport vector machines and neural network methodsand compare with Keras and SklearnBecause the data is incompletea potential factor model is also used to rebuild the lost data